ClusterCockpit/cc-backend
1
0
Fork 0
mirror of https://github.com/ClusterCockpit/cc-backend synced 2025-10-31 16:05:06 +01:00
Code Issues Packages Releases Activity

Merge branch 'master' into 37-provide-a-s3-compatible-storage-backend-for-the-job-archive

Browse Source
This commit is contained in:
Jan Eitzinger
2025-10-23 15:23:32 +02:00
parent 256d1b85f6 1a16851ad0
commit a371db5623
294 changed files with 35907 additions and 19676 deletions
Download Patch File Download Diff File Expand all files Collapse all files

203
pkg/archive/archive.go
View File

@@ -1,26 +1,20 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
package archive
import (
"bufio"
"bytes"
"compress/gzip"
"encoding/json"
"fmt"
"io"
"path/filepath"
"strconv"
"sync"
"github.com/ClusterCockpit/cc-backend/internal/config"
"github.com/ClusterCockpit/cc-backend/pkg/log"
"github.com/ClusterCockpit/cc-backend/pkg/lrucache"
"github.com/ClusterCockpit/cc-backend/pkg/schema"
)
const Version uint64 = 1
const Version uint64 = 2
type ArchiveBackend interface {
Init(rawConfig json.RawMessage) (uint64, error)
@@ -33,6 +27,8 @@ type ArchiveBackend interface {
LoadJobData(job *schema.Job) (schema.JobData, error)
LoadJobStats(job *schema.Job) (schema.ScopedJobStats, error)
LoadClusterCfg(name string) (*schema.Cluster, error)
StoreJobMeta(jobMeta *schema.JobMeta) error
@@ -60,105 +56,55 @@ type JobContainer struct {
}
var (
initOnce sync.Once
cache *lrucache.Cache = lrucache.New(128 * 1024 * 1024)
ar ArchiveBackend
useArchive bool
)
func getPath(
job *schema.Job,
rootPath string,
file string,
) string {
return filepath.Join(
getDirectory(job, rootPath), file)
}
func getDirectory(
job *schema.Job,
rootPath string,
) string {
lvl1, lvl2 := fmt.Sprintf("%d", job.JobID/1000), fmt.Sprintf("%03d", job.JobID%1000)
return filepath.Join(
rootPath,
job.Cluster,
lvl1, lvl2,
strconv.FormatInt(job.StartTime.Unix(), 10))
}
func loadJobMeta(b []byte) (*schema.JobMeta, error) {
if config.Keys.Validate {
if err := schema.Validate(schema.Meta, bytes.NewReader(b)); err != nil {
return &schema.JobMeta{}, fmt.Errorf("validate job meta: %v", err)
}
}
return DecodeJobMeta(bytes.NewReader(b))
}
func loadJobData(f io.Reader, key string, isCompressed bool) (schema.JobData, error) {
if isCompressed {
r, err := gzip.NewReader(f)
if err != nil {
log.Errorf(" %v", err)
return nil, err
}
defer r.Close()
if config.Keys.Validate {
if err := schema.Validate(schema.Data, r); err != nil {
return schema.JobData{}, fmt.Errorf("validate job data: %v", err)
}
}
return DecodeJobData(r, key)
} else {
if config.Keys.Validate {
if err := schema.Validate(schema.Data, bufio.NewReader(f)); err != nil {
return schema.JobData{}, fmt.Errorf("validate job data: %v", err)
}
}
return DecodeJobData(bufio.NewReader(f), key)
}
}
func Init(rawConfig json.RawMessage, disableArchive bool) error {
useArchive = !disableArchive
var err error
var cfg struct {
Kind string `json:"kind"`
}
initOnce.Do(func() {
useArchive = !disableArchive
if err := json.Unmarshal(rawConfig, &cfg); err != nil {
log.Warn("Error while unmarshaling raw config json")
return err
}
var cfg struct {
Kind string `json:"kind"`
}
switch cfg.Kind {
case "file":
ar = &FsArchive{}
// case "s3":
// ar = &S3Archive{}
default:
return fmt.Errorf("ARCHIVE/ARCHIVE > unkown archive backend '%s''", cfg.Kind)
}
if err = json.Unmarshal(rawConfig, &cfg); err != nil {
log.Warn("Error while unmarshaling raw config json")
return
}
version, err := ar.Init(rawConfig)
if err != nil {
log.Error("Error while initializing archiveBackend")
return err
}
log.Infof("Load archive version %d", version)
switch cfg.Kind {
case "file":
ar = &FsArchive{}
// case "s3":
// ar = &S3Archive{}
default:
err = fmt.Errorf("ARCHIVE/ARCHIVE > unkown archive backend '%s''", cfg.Kind)
}
return initClusterConfig()
var version uint64
version, err = ar.Init(rawConfig)
if err != nil {
log.Errorf("Error while initializing archiveBackend: %s", err.Error())
return
}
log.Infof("Load archive version %d", version)
err = initClusterConfig()
})
return err
}
func GetHandle() ArchiveBackend {
return ar
}
// Helper to metricdata.LoadAverages().
// Helper to metricdataloader.LoadAverages().
func LoadAveragesFromArchive(
job *schema.Job,
metrics []string,
@@ -166,7 +112,7 @@ func LoadAveragesFromArchive(
) error {
metaFile, err := ar.LoadJobMeta(job)
if err != nil {
log.Warn("Error while loading job metadata from archiveBackend")
log.Errorf("Error while loading job metadata from archiveBackend: %s", err.Error())
return err
}
@@ -181,16 +127,80 @@ func LoadAveragesFromArchive(
return nil
}
// Helper to metricdataloader.LoadJobStats().
func LoadStatsFromArchive(
job *schema.Job,
metrics []string,
) (map[string]schema.MetricStatistics, error) {
data := make(map[string]schema.MetricStatistics, len(metrics))
metaFile, err := ar.LoadJobMeta(job)
if err != nil {
log.Errorf("Error while loading job metadata from archiveBackend: %s", err.Error())
return data, err
}
for _, m := range metrics {
stat, ok := metaFile.Statistics[m]
if !ok {
data[m] = schema.MetricStatistics{Min: 0.0, Avg: 0.0, Max: 0.0}
continue
}
data[m] = schema.MetricStatistics{
Avg: stat.Avg,
Min: stat.Min,
Max: stat.Max,
}
}
return data, nil
}
// Helper to metricdataloader.LoadScopedJobStats().
func LoadScopedStatsFromArchive(
job *schema.Job,
metrics []string,
scopes []schema.MetricScope,
) (schema.ScopedJobStats, error) {
data, err := ar.LoadJobStats(job)
if err != nil {
log.Errorf("Error while loading job stats from archiveBackend: %s", err.Error())
return nil, err
}
return data, nil
}
func GetStatistics(job *schema.Job) (map[string]schema.JobStatistics, error) {
metaFile, err := ar.LoadJobMeta(job)
if err != nil {
log.Warn("Error while loading job metadata from archiveBackend")
log.Errorf("Error while loading job metadata from archiveBackend: %s", err.Error())
return nil, err
}
return metaFile.Statistics, nil
}
// If the job is archived, find its `meta.json` file and override the Metadata
// in that JSON file. If the job is not archived, nothing is done.
func UpdateMetadata(job *schema.Job, metadata map[string]string) error {
if job.State == schema.JobStateRunning || !useArchive {
return nil
}
jobMeta, err := ar.LoadJobMeta(job)
if err != nil {
log.Errorf("Error while loading job metadata from archiveBackend: %s", err.Error())
return err
}
for k, v := range metadata {
jobMeta.MetaData[k] = v
}
return ar.StoreJobMeta(jobMeta)
}
// If the job is archived, find its `meta.json` file and override the tags list
// in that JSON file. If the job is not archived, nothing is done.
func UpdateTags(job *schema.Job, tags []*schema.Tag) error {
@@ -200,15 +210,16 @@ func UpdateTags(job *schema.Job, tags []*schema.Tag) error {
jobMeta, err := ar.LoadJobMeta(job)
if err != nil {
log.Warn("Error while loading job metadata from archiveBackend")
log.Errorf("Error while loading job metadata from archiveBackend: %s", err.Error())
return err
}
jobMeta.Tags = make([]*schema.Tag, 0)
for _, tag := range tags {
jobMeta.Tags = append(jobMeta.Tags, &schema.Tag{
Name: tag.Name,
Type: tag.Type,
Name: tag.Name,
Type: tag.Type,
Scope: tag.Scope,
})
}

2
pkg/archive/archive_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2023 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

109
pkg/archive/clusterConfig.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -12,13 +12,16 @@ import (
"github.com/ClusterCockpit/cc-backend/pkg/schema"
)
var Clusters []*schema.Cluster
var nodeLists map[string]map[string]NodeList
var (
Clusters []*schema.Cluster
GlobalMetricList []*schema.GlobalMetricListItem
NodeLists map[string]map[string]NodeList
)
func initClusterConfig() error {
Clusters = []*schema.Cluster{}
nodeLists = map[string]map[string]NodeList{}
NodeLists = map[string]map[string]NodeList{}
metricLookup := make(map[string]schema.GlobalMetricListItem)
for _, c := range ar.GetClusters() {
@@ -49,11 +52,79 @@ func initClusterConfig() error {
if !mc.Scope.Valid() {
return errors.New("cluster.metricConfig.scope must be a valid scope ('node', 'scocket', ...)")
}
ml, ok := metricLookup[mc.Name]
if !ok {
metricLookup[mc.Name] = schema.GlobalMetricListItem{
Name: mc.Name, Scope: mc.Scope, Unit: mc.Unit, Footprint: mc.Footprint,
}
ml = metricLookup[mc.Name]
}
availability := schema.ClusterSupport{Cluster: cluster.Name}
scLookup := make(map[string]*schema.SubClusterConfig)
for _, scc := range mc.SubClusters {
scLookup[scc.Name] = scc
}
for _, sc := range cluster.SubClusters {
newMetric := &schema.MetricConfig{
Unit: mc.Unit,
Energy: mc.Energy,
Name: mc.Name,
Scope: mc.Scope,
Aggregation: mc.Aggregation,
Peak: mc.Peak,
Caution: mc.Caution,
Alert: mc.Alert,
Timestep: mc.Timestep,
Normal: mc.Normal,
LowerIsBetter: mc.LowerIsBetter,
}
if mc.Footprint != "" {
newMetric.Footprint = mc.Footprint
}
if cfg, ok := scLookup[sc.Name]; ok {
if !cfg.Remove {
availability.SubClusters = append(availability.SubClusters, sc.Name)
newMetric.Peak = cfg.Peak
newMetric.Normal = cfg.Normal
newMetric.Caution = cfg.Caution
newMetric.Alert = cfg.Alert
newMetric.Footprint = cfg.Footprint
newMetric.Energy = cfg.Energy
newMetric.LowerIsBetter = cfg.LowerIsBetter
sc.MetricConfig = append(sc.MetricConfig, *newMetric)
if newMetric.Footprint != "" {
sc.Footprint = append(sc.Footprint, newMetric.Name)
ml.Footprint = newMetric.Footprint
}
if newMetric.Energy != "" {
sc.EnergyFootprint = append(sc.EnergyFootprint, newMetric.Name)
}
}
} else {
availability.SubClusters = append(availability.SubClusters, sc.Name)
sc.MetricConfig = append(sc.MetricConfig, *newMetric)
if newMetric.Footprint != "" {
sc.Footprint = append(sc.Footprint, newMetric.Name)
}
if newMetric.Energy != "" {
sc.EnergyFootprint = append(sc.EnergyFootprint, newMetric.Name)
}
}
}
ml.Availability = append(metricLookup[mc.Name].Availability, availability)
metricLookup[mc.Name] = ml
}
Clusters = append(Clusters, cluster)
nodeLists[cluster.Name] = make(map[string]NodeList)
NodeLists[cluster.Name] = make(map[string]NodeList)
for _, sc := range cluster.SubClusters {
if sc.Nodes == "*" {
continue
@@ -63,15 +134,18 @@ func initClusterConfig() error {
if err != nil {
return fmt.Errorf("ARCHIVE/CLUSTERCONFIG > in %s/cluster.json: %w", cluster.Name, err)
}
nodeLists[cluster.Name][sc.Name] = nl
NodeLists[cluster.Name][sc.Name] = nl
}
}
for _, ml := range metricLookup {
GlobalMetricList = append(GlobalMetricList, &ml)
}
return nil
}
func GetCluster(cluster string) *schema.Cluster {
for _, c := range Clusters {
if c.Name == cluster {
return c
@@ -90,11 +164,10 @@ func GetSubCluster(cluster, subcluster string) (*schema.SubCluster, error) {
}
}
}
return nil, fmt.Errorf("Subcluster '%v' not found for cluster '%v', or cluster '%v' not configured!", subcluster, cluster, cluster)
return nil, fmt.Errorf("subcluster '%v' not found for cluster '%v', or cluster '%v' not configured", subcluster, cluster, cluster)
}
func GetMetricConfig(cluster, metric string) *schema.MetricConfig {
for _, c := range Clusters {
if c.Name == cluster {
for _, m := range c.MetricConfig {
@@ -110,7 +183,6 @@ func GetMetricConfig(cluster, metric string) *schema.MetricConfig {
// AssignSubCluster sets the `job.subcluster` property of the job based
// on its cluster and resources.
func AssignSubCluster(job *schema.BaseJob) error {
cluster := GetCluster(job.Cluster)
if cluster == nil {
return fmt.Errorf("ARCHIVE/CLUSTERCONFIG > unkown cluster: %v", job.Cluster)
@@ -130,7 +202,7 @@ func AssignSubCluster(job *schema.BaseJob) error {
}
host0 := job.Resources[0].Hostname
for sc, nl := range nodeLists[job.Cluster] {
for sc, nl := range NodeLists[job.Cluster] {
if nl != nil && nl.Contains(host0) {
job.SubCluster = sc
return nil
@@ -146,8 +218,7 @@ func AssignSubCluster(job *schema.BaseJob) error {
}
func GetSubClusterByNode(cluster, hostname string) (string, error) {
for sc, nl := range nodeLists[cluster] {
for sc, nl := range NodeLists[cluster] {
if nl != nil && nl.Contains(hostname) {
return sc, nil
}
@@ -164,3 +235,13 @@ func GetSubClusterByNode(cluster, hostname string) (string, error) {
return "", fmt.Errorf("ARCHIVE/CLUSTERCONFIG > no subcluster found for cluster %v and host %v", cluster, hostname)
}
func MetricIndex(mc []schema.MetricConfig, name string) (int, error) {
for i, m := range mc {
if m.Name == name {
return i, nil
}
}
return 0, fmt.Errorf("unknown metric name %s", name)
}

39
pkg/archive/clusterConfig_test.go Normal file
View File

@@ -0,0 +1,39 @@
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
package archive_test
import (
"encoding/json"
"testing"
"github.com/ClusterCockpit/cc-backend/pkg/archive"
)
func TestClusterConfig(t *testing.T) {
if err := archive.Init(json.RawMessage("{\"kind\": \"file\",\"path\": \"testdata/archive\"}"), false); err != nil {
t.Fatal(err)
}
sc, err := archive.GetSubCluster("fritz", "spr1tb")
if err != nil {
t.Fatal(err)
}
// spew.Dump(sc.MetricConfig)
if len(sc.Footprint) != 3 {
t.Fail()
}
if len(sc.MetricConfig) != 15 {
t.Fail()
}
for _, metric := range sc.MetricConfig {
if metric.LowerIsBetter && metric.Name != "mem_used" {
t.Fail()
}
}
// spew.Dump(archive.GlobalMetricList)
// t.Fail()
}

117
pkg/archive/fsBackend.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -40,6 +40,109 @@ type clusterInfo struct {
diskSize float64
}
func getDirectory(
job *schema.Job,
rootPath string,
) string {
lvl1, lvl2 := fmt.Sprintf("%d", job.JobID/1000), fmt.Sprintf("%03d", job.JobID%1000)
return filepath.Join(
rootPath,
job.Cluster,
lvl1, lvl2,
strconv.FormatInt(job.StartTime.Unix(), 10))
}
func getPath(
job *schema.Job,
rootPath string,
file string,
) string {
return filepath.Join(
getDirectory(job, rootPath), file)
}
func loadJobMeta(filename string) (*schema.JobMeta, error) {
b, err := os.ReadFile(filename)
if err != nil {
log.Errorf("loadJobMeta() > open file error: %v", err)
return &schema.JobMeta{}, err
}
if config.Keys.Validate {
if err := schema.Validate(schema.Meta, bytes.NewReader(b)); err != nil {
return &schema.JobMeta{}, fmt.Errorf("validate job meta: %v", err)
}
}
return DecodeJobMeta(bytes.NewReader(b))
}
func loadJobData(filename string, isCompressed bool) (schema.JobData, error) {
f, err := os.Open(filename)
if err != nil {
log.Errorf("fsBackend LoadJobData()- %v", err)
return nil, err
}
defer f.Close()
if isCompressed {
r, err := gzip.NewReader(f)
if err != nil {
log.Errorf(" %v", err)
return nil, err
}
defer r.Close()
if config.Keys.Validate {
if err := schema.Validate(schema.Data, r); err != nil {
return schema.JobData{}, fmt.Errorf("validate job data: %v", err)
}
}
return DecodeJobData(r, filename)
} else {
if config.Keys.Validate {
if err := schema.Validate(schema.Data, bufio.NewReader(f)); err != nil {
return schema.JobData{}, fmt.Errorf("validate job data: %v", err)
}
}
return DecodeJobData(bufio.NewReader(f), filename)
}
}
func loadJobStats(filename string, isCompressed bool) (schema.ScopedJobStats, error) {
f, err := os.Open(filename)
if err != nil {
log.Errorf("fsBackend LoadJobStats()- %v", err)
return nil, err
}
defer f.Close()
if isCompressed {
r, err := gzip.NewReader(f)
if err != nil {
log.Errorf(" %v", err)
return nil, err
}
defer r.Close()
if config.Keys.Validate {
if err := schema.Validate(schema.Data, r); err != nil {
return nil, fmt.Errorf("validate job data: %v", err)
}
}
return DecodeJobStats(r, filename)
} else {
if config.Keys.Validate {
if err := schema.Validate(schema.Data, bufio.NewReader(f)); err != nil {
return nil, fmt.Errorf("validate job data: %v", err)
}
}
return DecodeJobStats(bufio.NewReader(f), filename)
}
}
func (fsa *FsArchive) Init(rawConfig json.RawMessage) (uint64, error) {
var config FsArchiveConfig
if err := json.Unmarshal(rawConfig, &config); err != nil {
@@ -317,6 +420,18 @@ func (fsa *FsArchive) LoadJobData(job *schema.Job) (schema.JobData, error) {
return loadJobData(f, filename, isCompressed)
}
func (fsa *FsArchive) LoadJobStats(job *schema.Job) (schema.ScopedJobStats, error) {
var isCompressed bool = true
filename := getPath(job, fsa.path, "data.json.gz")
if !util.CheckFileExists(filename) {
filename = getPath(job, fsa.path, "data.json")
isCompressed = false
}
return loadJobStats(filename, isCompressed)
}
func (fsa *FsArchive) LoadJobMeta(job *schema.Job) (*schema.JobMeta, error) {
filename := getPath(job, fsa.path, "meta.json")
b, err := os.ReadFile(filename)

6
pkg/archive/fsBackend_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -49,10 +49,10 @@ func TestInit(t *testing.T) {
if fsa.path != "testdata/archive" {
t.Fail()
}
if version != 1 {
if version != 2 {
t.Fail()
}
if len(fsa.clusters) != 1 || fsa.clusters[0] != "emmy" {
if len(fsa.clusters) != 3 || fsa.clusters[1] != "emmy" {
t.Fail()
}
}

41
pkg/archive/json.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -9,8 +9,8 @@ import (
"io"
"time"
"github.com/ClusterCockpit/cc-backend/pkg/schema"
"github.com/ClusterCockpit/cc-backend/pkg/log"
"github.com/ClusterCockpit/cc-backend/pkg/schema"
)
func DecodeJobData(r io.Reader, k string) (schema.JobData, error) {
@@ -32,6 +32,43 @@ func DecodeJobData(r io.Reader, k string) (schema.JobData, error) {
return data.(schema.JobData), nil
}
func DecodeJobStats(r io.Reader, k string) (schema.ScopedJobStats, error) {
jobData, err := DecodeJobData(r, k)
// Convert schema.JobData to schema.ScopedJobStats
if jobData != nil {
scopedJobStats := make(schema.ScopedJobStats)
for metric, metricData := range jobData {
if _, ok := scopedJobStats[metric]; !ok {
scopedJobStats[metric] = make(map[schema.MetricScope][]*schema.ScopedStats)
}
for scope, jobMetric := range metricData {
if _, ok := scopedJobStats[metric][scope]; !ok {
scopedJobStats[metric][scope] = make([]*schema.ScopedStats, 0)
}
for _, series := range jobMetric.Series {
scopedJobStats[metric][scope] = append(scopedJobStats[metric][scope], &schema.ScopedStats{
Hostname: series.Hostname,
Id: series.Id,
Data: &series.Statistics,
})
}
// So that one can later check len(scopedJobStats[metric][scope]): Remove from map if empty
if len(scopedJobStats[metric][scope]) == 0 {
delete(scopedJobStats[metric], scope)
if len(scopedJobStats[metric]) == 0 {
delete(scopedJobStats, metric)
}
}
}
}
return scopedJobStats, nil
}
return nil, err
}
func DecodeJobMeta(r io.Reader) (*schema.JobMeta, error) {
var d schema.JobMeta
if err := json.NewDecoder(r).Decode(&d); err != nil {

2
pkg/archive/nodelist.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2
pkg/archive/nodelist_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2
pkg/archive/s3Backend.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2023 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2772
pkg/archive/testdata/archive/alex/cluster.json vendored Normal file
View File

File diff suppressed because it is too large Load Diff

1601
pkg/archive/testdata/archive/fritz/cluster.json vendored
View File

File diff suppressed because it is too large Load Diff

2
pkg/archive/testdata/archive/version.txt vendored
View File

@@ -1 +1 @@
1
2

238
pkg/log/log.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -46,12 +46,12 @@ var loglevel string = "info"
/* CONFIG */
func Init(lvl string, logdate bool) {
// Discard I/O for all writers below selected loglevel; <CRITICAL> is always written.
switch lvl {
case "crit":
ErrWriter = io.Discard
fallthrough
case "err", "fatal":
case "err":
WarnWriter = io.Discard
fallthrough
case "warn":
@@ -63,8 +63,7 @@ func Init(lvl string, logdate bool) {
// Nothing to do...
break
default:
fmt.Printf("pkg/log: Flag 'loglevel' has invalid value %#v\npkg/log: Will use default loglevel 'debug'\n", lvl)
//SetLogLevel("debug")
fmt.Printf("pkg/log: Flag 'loglevel' has invalid value %#v\npkg/log: Will use default loglevel '%s'\n", lvl, loglevel)
}
if !logdate {
@@ -84,109 +83,138 @@ func Init(lvl string, logdate bool) {
loglevel = lvl
}
/* PRINT */
// Private helper
func printStr(v ...interface{}) string {
return fmt.Sprint(v...)
}
// Uses Info() -> If errorpath required at some point:
// Will need own writer with 'Output(2, out)' to correctly render path
func Print(v ...interface{}) {
Info(v...)
}
func Debug(v ...interface{}) {
DebugLog.Output(2, printStr(v...))
}
func Info(v ...interface{}) {
InfoLog.Output(2, printStr(v...))
}
func Warn(v ...interface{}) {
WarnLog.Output(2, printStr(v...))
}
func Error(v ...interface{}) {
ErrLog.Output(2, printStr(v...))
}
// Writes panic stacktrace, but keeps application alive
func Panic(v ...interface{}) {
ErrLog.Output(2, printStr(v...))
panic("Panic triggered ...")
}
func Crit(v ...interface{}) {
CritLog.Output(2, printStr(v...))
}
// Writes critical log, stops application
func Fatal(v ...interface{}) {
CritLog.Output(2, printStr(v...))
os.Exit(1)
}
/* PRINT FORMAT*/
// Private helper
func printfStr(format string, v ...interface{}) string {
return fmt.Sprintf(format, v...)
}
// Uses Infof() -> If errorpath required at some point:
// Will need own writer with 'Output(2, out)' to correctly render path
func Printf(format string, v ...interface{}) {
Infof(format, v...)
}
func Debugf(format string, v ...interface{}) {
DebugLog.Output(2, printfStr(format, v...))
}
func Infof(format string, v ...interface{}) {
InfoLog.Output(2, printfStr(format, v...))
}
func Warnf(format string, v ...interface{}) {
WarnLog.Output(2, printfStr(format, v...))
}
func Errorf(format string, v ...interface{}) {
ErrLog.Output(2, printfStr(format, v...))
}
// Writes panic stacktrace, but keeps application alive
func Panicf(format string, v ...interface{}) {
ErrLog.Output(2, printfStr(format, v...))
panic("Panic triggered ...")
}
func Critf(format string, v ...interface{}) {
CritLog.Output(2, printfStr(format, v...))
}
// Writes crit log, stops application
func Fatalf(format string, v ...interface{}) {
CritLog.Output(2, printfStr(format, v...))
os.Exit(1)
}
/* HELPER */
func Loglevel() string {
return loglevel
}
/* SPECIAL */
/* PRIVATE HELPER */
// func Finfof(w io.Writer, format string, v ...interface{}) {
// if w != io.Discard {
// if logDateTime {
// currentTime := time.Now()
// fmt.Fprintf(InfoWriter, currentTime.String()+InfoPrefix+format+"\n", v...)
// } else {
// fmt.Fprintf(InfoWriter, InfoPrefix+format+"\n", v...)
// }
// }
// }
// Return unformatted string
func printStr(v ...interface{}) string {
return fmt.Sprint(v...)
}
// Return formatted string
func printfStr(format string, v ...interface{}) string {
return fmt.Sprintf(format, v...)
}
/* PRINT */
// Prints to STDOUT without string formatting; application continues.
// Used for special cases not requiring log information like date or location.
func Print(v ...interface{}) {
fmt.Fprintln(os.Stdout, v...)
}
// Prints to STDOUT without string formatting; application exits with error code 0.
// Used for exiting succesfully with message after expected outcome, e.g. successful single-call application runs.
func Exit(v ...interface{}) {
fmt.Fprintln(os.Stdout, v...)
os.Exit(0)
}
// Prints to STDOUT without string formatting; application exits with error code 1.
// Used for terminating with message after to be expected errors, e.g. wrong arguments or during init().
func Abort(v ...interface{}) {
fmt.Fprintln(os.Stdout, v...)
os.Exit(1)
}
// Prints to DEBUG writer without string formatting; application continues.
// Used for logging additional information, primarily for development.
func Debug(v ...interface{}) {
DebugLog.Output(2, printStr(v...))
}
// Prints to INFO writer without string formatting; application continues.
// Used for logging additional information, e.g. notable returns or common fail-cases.
func Info(v ...interface{}) {
InfoLog.Output(2, printStr(v...))
}
// Prints to WARNING writer without string formatting; application continues.
// Used for logging important information, e.g. uncommon edge-cases or administration related information.
func Warn(v ...interface{}) {
WarnLog.Output(2, printStr(v...))
}
// Prints to ERROR writer without string formatting; application continues.
// Used for logging errors, but code still can return default(s) or nil.
func Error(v ...interface{}) {
ErrLog.Output(2, printStr(v...))
}
// Prints to CRITICAL writer without string formatting; application exits with error code 1.
// Used for terminating on unexpected errors with date and code location.
func Fatal(v ...interface{}) {
CritLog.Output(2, printStr(v...))
os.Exit(1)
}
// Prints to PANIC function without string formatting; application exits with panic.
// Used for terminating on unexpected errors with stacktrace.
func Panic(v ...interface{}) {
panic(printStr(v...))
}
/* PRINT FORMAT*/
// Prints to STDOUT with string formatting; application continues.
// Used for special cases not requiring log information like date or location.
func Printf(format string, v ...interface{}) {
fmt.Fprintf(os.Stdout, format, v...)
}
// Prints to STDOUT with string formatting; application exits with error code 0.
// Used for exiting succesfully with message after expected outcome, e.g. successful single-call application runs.
func Exitf(format string, v ...interface{}) {
fmt.Fprintf(os.Stdout, format, v...)
os.Exit(0)
}
// Prints to STDOUT with string formatting; application exits with error code 1.
// Used for terminating with message after to be expected errors, e.g. wrong arguments or during init().
func Abortf(format string, v ...interface{}) {
fmt.Fprintf(os.Stdout, format, v...)
os.Exit(1)
}
// Prints to DEBUG writer with string formatting; application continues.
// Used for logging additional information, primarily for development.
func Debugf(format string, v ...interface{}) {
DebugLog.Output(2, printfStr(format, v...))
}
// Prints to INFO writer with string formatting; application continues.
// Used for logging additional information, e.g. notable returns or common fail-cases.
func Infof(format string, v ...interface{}) {
InfoLog.Output(2, printfStr(format, v...))
}
// Prints to WARNING writer with string formatting; application continues.
// Used for logging important information, e.g. uncommon edge-cases or administration related information.
func Warnf(format string, v ...interface{}) {
WarnLog.Output(2, printfStr(format, v...))
}
// Prints to ERROR writer with string formatting; application continues.
// Used for logging errors, but code still can return default(s) or nil.
func Errorf(format string, v ...interface{}) {
ErrLog.Output(2, printfStr(format, v...))
}
// Prints to CRITICAL writer with string formatting; application exits with error code 1.
// Used for terminating on unexpected errors with date and code location.
func Fatalf(format string, v ...interface{}) {
CritLog.Output(2, printfStr(format, v...))
os.Exit(1)
}
// Prints to PANIC function with string formatting; application exits with panic.
// Used for terminating on unexpected errors with stacktrace.
func Panicf(format string, v ...interface{}) {
panic(printfStr(format, v...))
}

2
pkg/lrucache/cache.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2
pkg/lrucache/cache_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2
pkg/lrucache/handler.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2
pkg/lrucache/handler_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

123
pkg/resampler/resampler.go Normal file
View File

@@ -0,0 +1,123 @@
package resampler
import (
"errors"
"fmt"
"math"
"github.com/ClusterCockpit/cc-backend/pkg/schema"
)
func SimpleResampler(data []schema.Float, old_frequency int64, new_frequency int64) ([]schema.Float, int64, error) {
if old_frequency == 0 || new_frequency == 0 || new_frequency <= old_frequency {
return data, old_frequency, nil
}
if new_frequency%old_frequency != 0 {
return nil, 0, errors.New("new sampling frequency should be multiple of the old frequency")
}
var step int = int(new_frequency / old_frequency)
var new_data_length = len(data) / step
if new_data_length == 0 || len(data) < 100 || new_data_length >= len(data) {
return data, old_frequency, nil
}
new_data := make([]schema.Float, new_data_length)
for i := 0; i < new_data_length; i++ {
new_data[i] = data[i*step]
}
return new_data, new_frequency, nil
}
// Inspired by one of the algorithms from https://skemman.is/bitstream/1946/15343/3/SS_MSthesis.pdf
// Adapted from https://github.com/haoel/downsampling/blob/master/core/lttb.go
func LargestTriangleThreeBucket(data []schema.Float, old_frequency int, new_frequency int) ([]schema.Float, int, error) {
if old_frequency == 0 || new_frequency == 0 || new_frequency <= old_frequency {
return data, old_frequency, nil
}
if new_frequency%old_frequency != 0 {
return nil, 0, errors.New(fmt.Sprintf("new sampling frequency : %d should be multiple of the old frequency : %d", new_frequency, old_frequency))
}
var step int = int(new_frequency / old_frequency)
var new_data_length = len(data) / step
if new_data_length == 0 || len(data) < 100 || new_data_length >= len(data) {
return data, old_frequency, nil
}
new_data := make([]schema.Float, 0, new_data_length)
// Bucket size. Leave room for start and end data points
bucketSize := float64(len(data)-2) / float64(new_data_length-2)
new_data = append(new_data, data[0]) // Always add the first point
// We have 3 pointers represent for
// > bucketLow - the current bucket's beginning location
// > bucketMiddle - the current bucket's ending location,
// also the beginning location of next bucket
// > bucketHight - the next bucket's ending location.
bucketLow := 1
bucketMiddle := int(math.Floor(bucketSize)) + 1
var prevMaxAreaPoint int
for i := 0; i < new_data_length-2; i++ {
bucketHigh := int(math.Floor(float64(i+2)*bucketSize)) + 1
if bucketHigh >= len(data)-1 {
bucketHigh = len(data) - 2
}
// Calculate point average for next bucket (containing c)
avgPointX, avgPointY := calculateAverageDataPoint(data[bucketMiddle:bucketHigh+1], int64(bucketMiddle))
// Get the range for current bucket
currBucketStart := bucketLow
currBucketEnd := bucketMiddle
// Point a
pointX := prevMaxAreaPoint
pointY := data[prevMaxAreaPoint]
maxArea := -1.0
var maxAreaPoint int
flag_ := 0
for ; currBucketStart < currBucketEnd; currBucketStart++ {
area := calculateTriangleArea(schema.Float(pointX), pointY, avgPointX, avgPointY, schema.Float(currBucketStart), data[currBucketStart])
if area > maxArea {
maxArea = area
maxAreaPoint = currBucketStart
}
if math.IsNaN(float64(avgPointY)) {
flag_ = 1
}
}
if flag_ == 1 {
new_data = append(new_data, schema.NaN) // Pick this point from the bucket
} else {
new_data = append(new_data, data[maxAreaPoint]) // Pick this point from the bucket
}
prevMaxAreaPoint = maxAreaPoint // This MaxArea point is the next's prevMAxAreaPoint
//move to the next window
bucketLow = bucketMiddle
bucketMiddle = bucketHigh
}
new_data = append(new_data, data[len(data)-1]) // Always add last
return new_data, new_frequency, nil
}

35
pkg/resampler/util.go Normal file
View File

@@ -0,0 +1,35 @@
package resampler
import (
"math"
"github.com/ClusterCockpit/cc-backend/pkg/schema"
)
func calculateTriangleArea(paX, paY, pbX, pbY, pcX, pcY schema.Float) float64 {
area := ((paX-pcX)*(pbY-paY) - (paX-pbX)*(pcY-paY)) * 0.5
return math.Abs(float64(area))
}
func calculateAverageDataPoint(points []schema.Float, xStart int64) (avgX schema.Float, avgY schema.Float) {
flag := 0
for _, point := range points {
avgX += schema.Float(xStart)
avgY += point
xStart++
if math.IsNaN(float64(point)) {
flag = 1
}
}
l := schema.Float(len(points))
avgX /= l
avgY /= l
if flag == 1 {
return avgX, schema.NaN
} else {
return avgX, avgY
}
}

142
pkg/runtimeEnv/setup.go Normal file
View File

@@ -0,0 +1,142 @@
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
package runtimeEnv
import (
"bufio"
"errors"
"fmt"
"os"
"os/exec"
"os/user"
"strconv"
"strings"
"syscall"
"github.com/ClusterCockpit/cc-backend/pkg/log"
)
// Very simple and limited .env file reader.
// All variable definitions found are directly
// added to the processes environment.
func LoadEnv(file string) error {
f, err := os.Open(file)
if err != nil {
log.Error("Error while opening .env file")
return err
}
defer f.Close()
s := bufio.NewScanner(bufio.NewReader(f))
for s.Scan() {
line := s.Text()
if strings.HasPrefix(line, "#") || len(line) == 0 {
continue
}
if strings.Contains(line, "#") {
return errors.New("'#' are only supported at the start of a line")
}
line = strings.TrimPrefix(line, "export ")
parts := strings.SplitN(line, "=", 2)
if len(parts) != 2 {
return fmt.Errorf("RUNTIME/SETUP > unsupported line: %#v", line)
}
key := strings.TrimSpace(parts[0])
val := strings.TrimSpace(parts[1])
if strings.HasPrefix(val, "\"") {
if !strings.HasSuffix(val, "\"") {
return fmt.Errorf("RUNTIME/SETUP > unsupported line: %#v", line)
}
runes := []rune(val[1 : len(val)-1])
sb := strings.Builder{}
for i := 0; i < len(runes); i++ {
if runes[i] == '\\' {
i++
switch runes[i] {
case 'n':
sb.WriteRune('\n')
case 'r':
sb.WriteRune('\r')
case 't':
sb.WriteRune('\t')
case '"':
sb.WriteRune('"')
default:
return fmt.Errorf("RUNTIME/SETUP > unsupported escape sequence in quoted string: backslash %#v", runes[i])
}
continue
}
sb.WriteRune(runes[i])
}
val = sb.String()
}
os.Setenv(key, val)
}
return s.Err()
}
// Changes the processes user and group to that
// specified in the config.json. The go runtime
// takes care of all threads (and not only the calling one)
// executing the underlying systemcall.
func DropPrivileges(username string, group string) error {
if group != "" {
g, err := user.LookupGroup(group)
if err != nil {
log.Warn("Error while looking up group")
return err
}
gid, _ := strconv.Atoi(g.Gid)
if err := syscall.Setgid(gid); err != nil {
log.Warn("Error while setting gid")
return err
}
}
if username != "" {
u, err := user.Lookup(username)
if err != nil {
log.Warn("Error while looking up user")
return err
}
uid, _ := strconv.Atoi(u.Uid)
if err := syscall.Setuid(uid); err != nil {
log.Warn("Error while setting uid")
return err
}
}
return nil
}
// If started via systemd, inform systemd that we are running:
// https://www.freedesktop.org/software/systemd/man/sd_notify.html
func SystemdNotifiy(ready bool, status string) {
if os.Getenv("NOTIFY_SOCKET") == "" {
// Not started using systemd
return
}
args := []string{fmt.Sprintf("--pid=%d", os.Getpid())}
if ready {
args = append(args, "--ready")
}
if status != "" {
args = append(args, fmt.Sprintf("--status=%s", status))
}
cmd := exec.Command("systemd-notify", args...)
cmd.Run() // errors ignored on purpose, there is not much to do anyways.
}

132
pkg/schema/cluster.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -30,38 +30,47 @@ type MetricValue struct {
}
type SubCluster struct {
Name string `json:"name"`
Nodes string `json:"nodes"`
ProcessorType string `json:"processorType"`
SocketsPerNode int `json:"socketsPerNode"`
CoresPerSocket int `json:"coresPerSocket"`
ThreadsPerCore int `json:"threadsPerCore"`
FlopRateScalar MetricValue `json:"flopRateScalar"`
FlopRateSimd MetricValue `json:"flopRateSimd"`
MemoryBandwidth MetricValue `json:"memoryBandwidth"`
Topology Topology `json:"topology"`
Name string `json:"name"`
Nodes string `json:"nodes"`
ProcessorType string `json:"processorType"`
Topology Topology `json:"topology"`
FlopRateScalar MetricValue `json:"flopRateScalar"`
FlopRateSimd MetricValue `json:"flopRateSimd"`
MemoryBandwidth MetricValue `json:"memoryBandwidth"`
MetricConfig []MetricConfig `json:"metricConfig,omitempty"`
Footprint []string `json:"footprint,omitempty"`
EnergyFootprint []string `json:"energyFootprint,omitempty"`
SocketsPerNode int `json:"socketsPerNode"`
CoresPerSocket int `json:"coresPerSocket"`
ThreadsPerCore int `json:"threadsPerCore"`
}
type SubClusterConfig struct {
Name string `json:"name"`
Peak float64 `json:"peak"`
Normal float64 `json:"normal"`
Caution float64 `json:"caution"`
Alert float64 `json:"alert"`
Remove bool `json:"remove"`
Name string `json:"name"`
Footprint string `json:"footprint,omitempty"`
Energy string `json:"energy"`
Peak float64 `json:"peak"`
Normal float64 `json:"normal"`
Caution float64 `json:"caution"`
Alert float64 `json:"alert"`
Remove bool `json:"remove"`
LowerIsBetter bool `json:"lowerIsBetter"`
}
type MetricConfig struct {
Name string `json:"name"`
Unit Unit `json:"unit"`
Scope MetricScope `json:"scope"`
Aggregation string `json:"aggregation"`
Timestep int `json:"timestep"`
Peak float64 `json:"peak"`
Normal float64 `json:"normal"`
Caution float64 `json:"caution"`
Alert float64 `json:"alert"`
SubClusters []*SubClusterConfig `json:"subClusters,omitempty"`
Unit Unit `json:"unit"`
Energy string `json:"energy"`
Name string `json:"name"`
Scope MetricScope `json:"scope"`
Aggregation string `json:"aggregation"`
Footprint string `json:"footprint,omitempty"`
SubClusters []*SubClusterConfig `json:"subClusters,omitempty"`
Peak float64 `json:"peak"`
Caution float64 `json:"caution"`
Alert float64 `json:"alert"`
Timestep int `json:"timestep"`
Normal float64 `json:"normal"`
LowerIsBetter bool `json:"lowerIsBetter"`
}
type Cluster struct {
@@ -70,14 +79,27 @@ type Cluster struct {
SubClusters []*SubCluster `json:"subClusters"`
}
type ClusterSupport struct {
Cluster string `json:"cluster"`
SubClusters []string `json:"subclusters"`
}
type GlobalMetricListItem struct {
Name string `json:"name"`
Unit Unit `json:"unit"`
Scope MetricScope `json:"scope"`
Footprint string `json:"footprint,omitempty"`
Availability []ClusterSupport `json:"availability"`
}
// Return a list of socket IDs given a list of hwthread IDs. Even if just one
// hwthread is in that socket, add it to the list. If no hwthreads other than
// those in the argument list are assigned to one of the sockets in the first
// return value, return true as the second value. TODO: Optimize this, there
// must be a more efficient way/algorithm.
func (topo *Topology) GetSocketsFromHWThreads(
hwthreads []int) (sockets []int, exclusive bool) {
hwthreads []int,
) (sockets []int, exclusive bool) {
socketsMap := map[int]int{}
for _, hwthread := range hwthreads {
for socket, hwthreadsInSocket := range topo.Socket {
@@ -100,14 +122,46 @@ func (topo *Topology) GetSocketsFromHWThreads(
return sockets, exclusive
}
// Return a list of socket IDs given a list of core IDs. Even if just one
// core is in that socket, add it to the list. If no cores other than
// those in the argument list are assigned to one of the sockets in the first
// return value, return true as the second value. TODO: Optimize this, there
// must be a more efficient way/algorithm.
func (topo *Topology) GetSocketsFromCores (
cores []int,
) (sockets []int, exclusive bool) {
socketsMap := map[int]int{}
for _, core := range cores {
for _, hwthreadInCore := range topo.Core[core] {
for socket, hwthreadsInSocket := range topo.Socket {
for _, hwthreadInSocket := range hwthreadsInSocket {
if hwthreadInCore == hwthreadInSocket {
socketsMap[socket] += 1
}
}
}
}
}
exclusive = true
hwthreadsPerSocket := len(topo.Node) / len(topo.Socket)
sockets = make([]int, 0, len(socketsMap))
for socket, count := range socketsMap {
sockets = append(sockets, socket)
exclusive = exclusive && count == hwthreadsPerSocket
}
return sockets, exclusive
}
// Return a list of core IDs given a list of hwthread IDs. Even if just one
// hwthread is in that core, add it to the list. If no hwthreads other than
// those in the argument list are assigned to one of the cores in the first
// return value, return true as the second value. TODO: Optimize this, there
// must be a more efficient way/algorithm.
func (topo *Topology) GetCoresFromHWThreads(
hwthreads []int) (cores []int, exclusive bool) {
hwthreads []int,
) (cores []int, exclusive bool) {
coresMap := map[int]int{}
for _, hwthread := range hwthreads {
for core, hwthreadsInCore := range topo.Core {
@@ -136,8 +190,8 @@ func (topo *Topology) GetCoresFromHWThreads(
// memory domains in the first return value, return true as the second value.
// TODO: Optimize this, there must be a more efficient way/algorithm.
func (topo *Topology) GetMemoryDomainsFromHWThreads(
hwthreads []int) (memDoms []int, exclusive bool) {
hwthreads []int,
) (memDoms []int, exclusive bool) {
memDomsMap := map[int]int{}
for _, hwthread := range hwthreads {
for memDom, hwthreadsInmemDom := range topo.MemoryDomain {
@@ -172,7 +226,17 @@ func (topo *Topology) GetAcceleratorID(id int) (string, error) {
}
}
func (topo *Topology) GetAcceleratorIDs() ([]int, error) {
// Return list of hardware (string) accelerator IDs
func (topo *Topology) GetAcceleratorIDs() []string {
accels := make([]string, 0)
for _, accel := range topo.Accelerators {
accels = append(accels, accel.ID)
}
return accels
}
// Outdated? Or: Return indices of accelerators in parent array?
func (topo *Topology) GetAcceleratorIDsAsInt() ([]int, error) {
accels := make([]int, 0)
for _, accel := range topo.Accelerators {
id, err := strconv.Atoi(accel.ID)

51
pkg/schema/config.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -23,6 +23,12 @@ type LdapConfig struct {
SyncUserOnLogin bool `json:"syncUserOnLogin"`
}
type OpenIDConfig struct {
Provider string `json:"provider"`
SyncUserOnLogin bool `json:"syncUserOnLogin"`
UpdateUserOnLogin bool `json:"updateUserOnLogin"`
}
type JWTAuthConfig struct {
// Specifies for how long a JWT token shall be valid
// as a string parsable by time.ParseDuration().
@@ -40,6 +46,9 @@ type JWTAuthConfig struct {
// Should an non-existent user be added to the DB based on the information in the token
SyncUserOnLogin bool `json:"syncUserOnLogin"`
// Should an existent user be updated in the DB based on the information in the token
UpdateUserOnLogin bool `json:"updateUserOnLogin"`
}
type IntRange struct {
@@ -48,8 +57,9 @@ type IntRange struct {
}
type TimeRange struct {
From *time.Time `json:"from"`
To *time.Time `json:"to"`
From *time.Time `json:"from"`
To *time.Time `json:"to"`
Range string `json:"range,omitempty"`
}
type FilterRanges struct {
@@ -65,10 +75,24 @@ type ClusterConfig struct {
}
type Retention struct {
Age int `json:"age"`
IncludeDB bool `json:"includeDB"`
Policy string `json:"policy"`
Location string `json:"location"`
Age int `json:"age"`
IncludeDB bool `json:"includeDB"`
}
type ResampleConfig struct {
// Array of resampling target resolutions, in seconds; Example: [600,300,60]
Resolutions []int `json:"resolutions"`
// Trigger next zoom level at less than this many visible datapoints
Trigger int `json:"trigger"`
}
type CronFrequency struct {
// Duration Update Worker [Defaults to '5m']
DurationWorker string `json:"duration-worker"`
// Metric-Footprint Update Worker [Defaults to '10m']
FootprintWorker string `json:"footprint-worker"`
}
// Format of the configuration (file). See below for the defaults.
@@ -76,7 +100,7 @@ type ProgramConfig struct {
// Address where the http (or https) server will listen on (for example: 'localhost:80').
Addr string `json:"addr"`
// Addresses from which secured API endpoints can be reached
// Addresses from which secured admin API endpoints can be reached, can be wildcard "*"
ApiAllowedIPs []string `json:"apiAllowedIPs"`
// Drop root permissions once .env was read and the port was taken.
@@ -109,8 +133,9 @@ type ProgramConfig struct {
Validate bool `json:"validate"`
// For LDAP Authentication and user synchronisation.
LdapConfig *LdapConfig `json:"ldap"`
JwtConfig *JWTAuthConfig `json:"jwts"`
LdapConfig *LdapConfig `json:"ldap"`
JwtConfig *JWTAuthConfig `json:"jwts"`
OpenIDConfig *OpenIDConfig `json:"oidc"`
// If 0 or empty, the session does not expire!
SessionMaxAge string `json:"session-max-age"`
@@ -127,6 +152,9 @@ type ProgramConfig struct {
// be provided! Most options here can be overwritten by the user.
UiDefaults map[string]interface{} `json:"ui-defaults"`
// If exists, will enable dynamic zoom in frontend metric plots using the configured values
EnableResampling *ResampleConfig `json:"enable-resampling"`
// Where to store MachineState files
MachineStateDir string `json:"machine-state-dir"`
@@ -136,6 +164,13 @@ type ProgramConfig struct {
// Defines time X in seconds in which jobs are considered to be "short" and will be filtered in specific views.
ShortRunningJobsDuration int `json:"short-running-jobs-duration"`
// Energy Mix CO2 Emission Constant [g/kWh]
// If entered, displays estimated CO2 emission for job based on jobs totalEnergy
EmissionConstant int `json:"emission-constant"`
// Frequency of cron job workers
CronFrequency *CronFrequency `json:"cron-frequency"`
// Array of Clusters
Clusters []*ClusterConfig `json:"clusters"`
}

2
pkg/schema/float.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

90
pkg/schema/job.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -16,30 +16,33 @@ import (
// Common subset of Job and JobMeta. Use one of those, not this type directly.
type BaseJob struct {
// The unique identifier of a job
JobID int64 `json:"jobId" db:"job_id" example:"123000"`
User string `json:"user" db:"user" example:"abcd100h"` // The unique identifier of a user
Project string `json:"project" db:"project" example:"abcd200"` // The unique identifier of a project
Cluster string `json:"cluster" db:"cluster" example:"fritz"` // The unique identifier of a cluster
SubCluster string `json:"subCluster" db:"subcluster" example:"main"` // The unique identifier of a sub cluster
Partition string `json:"partition,omitempty" db:"partition" example:"main"` // The Slurm partition to which the job was submitted
ArrayJobId int64 `json:"arrayJobId,omitempty" db:"array_job_id" example:"123000"` // The unique identifier of an array job
NumNodes int32 `json:"numNodes" db:"num_nodes" example:"2" minimum:"1"` // Number of nodes used (Min > 0)
// NumCores int32 `json:"numCores" db:"num_cores" example:"20" minimum:"1"` // Number of HWThreads used (Min > 0)
NumHWThreads int32 `json:"numHwthreads,omitempty" db:"num_hwthreads" example:"20" minimum:"1"` // Number of HWThreads used (Min > 0)
NumAcc int32 `json:"numAcc,omitempty" db:"num_acc" example:"2" minimum:"1"` // Number of accelerators used (Min > 0)
Exclusive int32 `json:"exclusive" db:"exclusive" example:"1" minimum:"0" maximum:"2"` // Specifies how nodes are shared: 0 - Shared among multiple jobs of multiple users, 1 - Job exclusive (Default), 2 - Shared among multiple jobs of same user
MonitoringStatus int32 `json:"monitoringStatus,omitempty" db:"monitoring_status" example:"1" minimum:"0" maximum:"3"` // State of monitoring system during job run: 0 - Disabled, 1 - Running or Archiving (Default), 2 - Archiving Failed, 3 - Archiving Successfull
SMT int32 `json:"smt,omitempty" db:"smt" example:"4"` // SMT threads used by job
State JobState `json:"jobState" db:"job_state" example:"completed" enums:"completed,failed,cancelled,stopped,timeout,out_of_memory"` // Final state of job
Duration int32 `json:"duration" db:"duration" example:"43200" minimum:"1"` // Duration of job in seconds (Min > 0)
Walltime int64 `json:"walltime,omitempty" db:"walltime" example:"86400" minimum:"1"` // Requested walltime of job in seconds (Min > 0)
Tags []*Tag `json:"tags,omitempty"` // List of tags
RawResources []byte `json:"-" db:"resources"` // Resources used by job [As Bytes]
Resources []*Resource `json:"resources"` // Resources used by job
RawMetaData []byte `json:"-" db:"meta_data"` // Additional information about the job [As Bytes]
MetaData map[string]string `json:"metaData"` // Additional information about the job
ConcurrentJobs JobLinkResultList `json:"concurrentJobs"`
Cluster string `json:"cluster" db:"cluster" example:"fritz"`
SubCluster string `json:"subCluster" db:"subcluster" example:"main"`
Partition string `json:"partition,omitempty" db:"cluster_partition" example:"main"`
Project string `json:"project" db:"project" example:"abcd200"`
User string `json:"user" db:"hpc_user" example:"abcd100h"`
State JobState `json:"jobState" db:"job_state" example:"completed" enums:"completed,failed,cancelled,stopped,timeout,out_of_memory"`
Tags []*Tag `json:"tags,omitempty"`
RawEnergyFootprint []byte `json:"-" db:"energy_footprint"`
RawFootprint []byte `json:"-" db:"footprint"`
RawMetaData []byte `json:"-" db:"meta_data"`
RawResources []byte `json:"-" db:"resources"`
Resources []*Resource `json:"resources"`
EnergyFootprint map[string]float64 `json:"energyFootprint"`
Footprint map[string]float64 `json:"footprint"`
MetaData map[string]string `json:"metaData"`
ConcurrentJobs JobLinkResultList `json:"concurrentJobs"`
Energy float64 `json:"energy" db:"energy"`
ArrayJobId int64 `json:"arrayJobId,omitempty" db:"array_job_id" example:"123000"`
Walltime int64 `json:"walltime,omitempty" db:"walltime" example:"86400" minimum:"1"`
JobID int64 `json:"jobId" db:"job_id" example:"123000"`
Duration int32 `json:"duration" db:"duration" example:"43200" minimum:"1"`
SMT int32 `json:"smt,omitempty" db:"smt" example:"4"`
MonitoringStatus int32 `json:"monitoringStatus,omitempty" db:"monitoring_status" example:"1" minimum:"0" maximum:"3"`
Exclusive int32 `json:"exclusive" db:"exclusive" example:"1" minimum:"0" maximum:"2"`
NumAcc int32 `json:"numAcc,omitempty" db:"num_acc" example:"2" minimum:"1"`
NumHWThreads int32 `json:"numHwthreads,omitempty" db:"num_hwthreads" example:"20" minimum:"1"`
NumNodes int32 `json:"numNodes" db:"num_nodes" example:"2" minimum:"1"`
}
// Job struct type
@@ -49,19 +52,10 @@ type BaseJob struct {
// Job model
// @Description Information of a HPC job.
type Job struct {
// The unique identifier of a job in the database
ID int64 `json:"id" db:"id"`
StartTime time.Time `json:"startTime"`
BaseJob
StartTimeUnix int64 `json:"-" db:"start_time" example:"1649723812"` // Start epoch time stamp in seconds
StartTime time.Time `json:"startTime"` // Start time as 'time.Time' data type
MemUsedMax float64 `json:"memUsedMax" db:"mem_used_max"` // MemUsedMax as Float64
FlopsAnyAvg float64 `json:"flopsAnyAvg" db:"flops_any_avg"` // FlopsAnyAvg as Float64
MemBwAvg float64 `json:"memBwAvg" db:"mem_bw_avg"` // MemBwAvg as Float64
LoadAvg float64 `json:"loadAvg" db:"load_avg"` // LoadAvg as Float64
NetBwAvg float64 `json:"-" db:"net_bw_avg"` // NetBwAvg as Float64
NetDataVolTotal float64 `json:"-" db:"net_data_vol_total"` // NetDataVolTotal as Float64
FileBwAvg float64 `json:"-" db:"file_bw_avg"` // FileBwAvg as Float64
FileDataVolTotal float64 `json:"-" db:"file_data_vol_total"` // FileDataVolTotal as Float64
ID int64 `json:"id" db:"id"`
StartTimeUnix int64 `json:"-" db:"start_time" example:"1649723812"`
}
// JobMeta struct type
@@ -88,11 +82,10 @@ type JobLinkResultList struct {
// JobMeta model
// @Description Meta data information of a HPC job.
type JobMeta struct {
// The unique identifier of a job in the database
ID *int64 `json:"id,omitempty"`
ID *int64 `json:"id,omitempty"`
Statistics map[string]JobStatistics `json:"statistics"`
BaseJob
StartTime int64 `json:"startTime" db:"start_time" example:"1649723812" minimum:"1"` // Start epoch time stamp in seconds (Min > 0)
Statistics map[string]JobStatistics `json:"statistics"` // Metric statistics of job
StartTime int64 `json:"startTime" db:"start_time" example:"1649723812" minimum:"1"`
}
const (
@@ -124,18 +117,19 @@ type JobStatistics struct {
// Tag model
// @Description Defines a tag using name and type.
type Tag struct {
ID int64 `json:"id" db:"id"` // The unique DB identifier of a tag
Type string `json:"type" db:"tag_type" example:"Debug"` // Tag Type
Name string `json:"name" db:"tag_name" example:"Testjob"` // Tag Name
Type string `json:"type" db:"tag_type" example:"Debug"`
Name string `json:"name" db:"tag_name" example:"Testjob"`
Scope string `json:"scope" db:"tag_scope" example:"global"`
ID int64 `json:"id" db:"id"`
}
// Resource model
// @Description A resource used by a job
type Resource struct {
Hostname string `json:"hostname"` // Name of the host (= node)
HWThreads []int `json:"hwthreads,omitempty"` // List of OS processor ids
Accelerators []string `json:"accelerators,omitempty"` // List of of accelerator device ids
Configuration string `json:"configuration,omitempty"` // The configuration options of the node
Hostname string `json:"hostname"`
Configuration string `json:"configuration,omitempty"`
HWThreads []int `json:"hwthreads,omitempty"`
Accelerators []string `json:"accelerators,omitempty"`
}
type JobState string

83
pkg/schema/metrics.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -10,22 +10,31 @@ import (
"math"
"sort"
"unsafe"
"github.com/ClusterCockpit/cc-backend/internal/util"
)
type JobData map[string]map[MetricScope]*JobMetric
type ScopedJobStats map[string]map[MetricScope][]*ScopedStats
type JobMetric struct {
Unit Unit `json:"unit"`
Timestep int `json:"timestep"`
Series []Series `json:"series"`
StatisticsSeries *StatsSeries `json:"statisticsSeries,omitempty"`
Unit Unit `json:"unit"`
Series []Series `json:"series"`
Timestep int `json:"timestep"`
}
type Series struct {
Hostname string `json:"hostname"`
Id *string `json:"id,omitempty"`
Statistics MetricStatistics `json:"statistics"`
Hostname string `json:"hostname"`
Data []Float `json:"data"`
Statistics MetricStatistics `json:"statistics"`
}
type ScopedStats struct {
Hostname string `json:"hostname"`
Id *string `json:"id,omitempty"`
Data *MetricStatistics `json:"data"`
}
type MetricStatistics struct {
@@ -35,10 +44,11 @@ type MetricStatistics struct {
}
type StatsSeries struct {
Percentiles map[int][]Float `json:"percentiles,omitempty"`
Mean []Float `json:"mean"`
Median []Float `json:"median"`
Min []Float `json:"min"`
Max []Float `json:"max"`
Percentiles map[int][]Float `json:"percentiles,omitempty"`
}
type MetricScope string
@@ -121,6 +131,7 @@ func (jd *JobData) Size() int {
if metric.StatisticsSeries != nil {
n += len(metric.StatisticsSeries.Max)
n += len(metric.StatisticsSeries.Mean)
n += len(metric.StatisticsSeries.Median)
n += len(metric.StatisticsSeries.Min)
}
@@ -149,53 +160,74 @@ func (jm *JobMetric) AddStatisticsSeries() {
}
}
min, mean, max := make([]Float, n), make([]Float, n), make([]Float, n)
// mean := make([]Float, n)
min, median, max := make([]Float, n), make([]Float, n), make([]Float, n)
i := 0
for ; i < m; i++ {
smin, ssum, smax := math.MaxFloat32, 0.0, -math.MaxFloat32
seriesCount := len(jm.Series)
// ssum := 0.0
smin, smed, smax := math.MaxFloat32, make([]float64, seriesCount), -math.MaxFloat32
notnan := 0
for j := 0; j < len(jm.Series); j++ {
for j := 0; j < seriesCount; j++ {
x := float64(jm.Series[j].Data[i])
if math.IsNaN(x) {
continue
}
notnan += 1
ssum += x
// ssum += x
smed[j] = x
smin = math.Min(smin, x)
smax = math.Max(smax, x)
}
if notnan < 3 {
min[i] = NaN
mean[i] = NaN
// mean[i] = NaN
median[i] = NaN
max[i] = NaN
} else {
min[i] = Float(smin)
mean[i] = Float(ssum / float64(notnan))
// mean[i] = Float(ssum / float64(notnan))
max[i] = Float(smax)
medianRaw, err := util.Median(smed)
if err != nil {
median[i] = NaN
} else {
median[i] = Float(medianRaw)
}
}
}
for ; i < n; i++ {
min[i] = NaN
mean[i] = NaN
// mean[i] = NaN
median[i] = NaN
max[i] = NaN
}
if smooth {
for i := 2; i < len(mean)-2; i++ {
for i := 2; i < len(median)-2; i++ {
if min[i].IsNaN() {
continue
}
min[i] = (min[i-2] + min[i-1] + min[i] + min[i+1] + min[i+2]) / 5
max[i] = (max[i-2] + max[i-1] + max[i] + max[i+1] + max[i+2]) / 5
mean[i] = (mean[i-2] + mean[i-1] + mean[i] + mean[i+1] + mean[i+2]) / 5
// mean[i] = (mean[i-2] + mean[i-1] + mean[i] + mean[i+1] + mean[i+2]) / 5
// Reduce Median further
smoothRaw := []float64{float64(median[i-2]), float64(median[i-1]), float64(median[i]), float64(median[i+1]), float64(median[i+2])}
smoothMedian, err := util.Median(smoothRaw)
if err != nil {
median[i] = NaN
} else {
median[i] = Float(smoothMedian)
}
}
}
jm.StatisticsSeries = &StatsSeries{Mean: mean, Min: min, Max: max}
jm.StatisticsSeries = &StatsSeries{Median: median, Min: min, Max: max} // Mean: mean
}
func (jd *JobData) AddNodeScope(metric string) bool {
@@ -204,7 +236,7 @@ func (jd *JobData) AddNodeScope(metric string) bool {
return false
}
var maxScope MetricScope = MetricScopeInvalid
maxScope := MetricScopeInvalid
for scope := range scopes {
maxScope = maxScope.Max(scope)
}
@@ -266,6 +298,21 @@ func (jd *JobData) AddNodeScope(metric string) bool {
return true
}
func (jd *JobData) RoundMetricStats() {
// TODO: Make Digit-Precision Configurable? (Currently: Fixed to 2 Digits)
for _, scopes := range *jd {
for _, jm := range scopes {
for index := range jm.Series {
jm.Series[index].Statistics = MetricStatistics{
Avg: (math.Round(jm.Series[index].Statistics.Avg*100) / 100),
Min: (math.Round(jm.Series[index].Statistics.Min*100) / 100),
Max: (math.Round(jm.Series[index].Statistics.Max*100) / 100),
}
}
}
}
}
func (jm *JobMetric) AddPercentiles(ps []int) bool {
if jm.StatisticsSeries == nil {
jm.AddStatisticsSeries()

617
pkg/schema/schemas/cluster.schema.json
View File

@@ -1,284 +1,339 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://cluster.schema.json",
"title": "HPC cluster description",
"description": "Meta data information of a HPC cluster",
"type": "object",
"properties": {
"name": {
"description": "The unique identifier of a cluster",
"type": "string"
},
"metricConfig": {
"description": "Metric specifications",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "Metric name",
"type": "string"
},
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"scope": {
"description": "Native measurement resolution",
"type": "string"
},
"timestep": {
"description": "Frequency of timeseries points",
"type": "integer"
},
"aggregation": {
"description": "How the metric is aggregated",
"type": "string",
"enum": [
"sum",
"avg"
]
},
"peak": {
"description": "Metric peak threshold (Upper metric limit)",
"type": "number"
},
"normal": {
"description": "Metric normal threshold",
"type": "number"
},
"caution": {
"description": "Metric caution threshold (Suspicious but does not require immediate action)",
"type": "number"
},
"alert": {
"description": "Metric alert threshold (Requires immediate action)",
"type": "number"
},
"subClusters": {
"description": "Array of cluster hardware partition metric thresholds",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "Hardware partition name",
"type": "string"
},
"peak": {
"type": "number"
},
"normal": {
"type": "number"
},
"caution": {
"type": "number"
},
"alert": {
"type": "number"
},
"remove": {
"type": "boolean"
}
},
"required": [
"name"
]
}
}
},
"required": [
"name",
"unit",
"scope",
"timestep",
"aggregation",
"peak",
"normal",
"caution",
"alert"
]
},
"minItems": 1
},
"subClusters": {
"description": "Array of cluster hardware partitions",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "Hardware partition name",
"type": "string"
},
"processorType": {
"description": "Processor type",
"type": "string"
},
"socketsPerNode": {
"description": "Number of sockets per node",
"type": "integer"
},
"coresPerSocket": {
"description": "Number of cores per socket",
"type": "integer"
},
"threadsPerCore": {
"description": "Number of SMT threads per core",
"type": "integer"
},
"flopRateScalar": {
"description": "Theoretical node peak flop rate for scalar code in GFlops/s",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"value": {
"type": "number"
}
}
},
"flopRateSimd": {
"description": "Theoretical node peak flop rate for SIMD code in GFlops/s",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"value": {
"type": "number"
}
}
},
"memoryBandwidth": {
"description": "Theoretical node peak memory bandwidth in GB/s",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"value": {
"type": "number"
}
}
},
"nodes": {
"description": "Node list expression",
"type": "string"
},
"topology": {
"description": "Node topology",
"type": "object",
"properties": {
"node": {
"description": "HwTread lists of node",
"type": "array",
"items": {
"type": "integer"
}
},
"socket": {
"description": "HwTread lists of sockets",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"memoryDomain": {
"description": "HwTread lists of memory domains",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"die": {
"description": "HwTread lists of dies",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"core": {
"description": "HwTread lists of cores",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"accelerators": {
"type": "array",
"description": "List of of accelerator devices",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "The unique device id"
},
"type": {
"type": "string",
"description": "The accelerator type",
"enum": [
"Nvidia GPU",
"AMD GPU",
"Intel GPU"
]
},
"model": {
"type": "string",
"description": "The accelerator model"
}
},
"required": [
"id",
"type",
"model"
]
}
}
},
"required": [
"node",
"socket",
"memoryDomain"
]
}
},
"required": [
"name",
"nodes",
"topology",
"processorType",
"socketsPerNode",
"coresPerSocket",
"threadsPerCore",
"flopRateScalar",
"flopRateSimd",
"memoryBandwidth"
]
},
"minItems": 1
}
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://cluster.schema.json",
"title": "HPC cluster description",
"description": "Meta data information of a HPC cluster",
"type": "object",
"properties": {
"name": {
"description": "The unique identifier of a cluster",
"type": "string"
},
"required": [
"name",
"metricConfig",
"subClusters"
]
"metricConfig": {
"description": "Metric specifications",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "Metric name",
"type": "string"
},
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"scope": {
"description": "Native measurement resolution",
"type": "string",
"enum": [
"node",
"socket",
"memoryDomain",
"core",
"hwthread",
"accelerator"
]
},
"timestep": {
"description": "Frequency of timeseries points in seconds",
"type": "integer"
},
"aggregation": {
"description": "How the metric is aggregated",
"type": "string",
"enum": [
"sum",
"avg"
]
},
"footprint": {
"description": "Is it a footprint metric and what type",
"type": "string",
"enum": [
"avg",
"max",
"min"
]
},
"energy": {
"description": "Is it used to calculate job energy",
"type": "string",
"enum": [
"power",
"energy"
]
},
"lowerIsBetter": {
"description": "Is lower better.",
"type": "boolean"
},
"peak": {
"description": "Metric peak threshold (Upper metric limit)",
"type": "number"
},
"normal": {
"description": "Metric normal threshold",
"type": "number"
},
"caution": {
"description": "Metric caution threshold (Suspicious but does not require immediate action)",
"type": "number"
},
"alert": {
"description": "Metric alert threshold (Requires immediate action)",
"type": "number"
},
"subClusters": {
"description": "Array of cluster hardware partition metric thresholds",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "Hardware partition name",
"type": "string"
},
"footprint": {
"description": "Is it a footprint metric and what type. Overwrite global setting",
"type": "string",
"enum": [
"avg",
"max",
"min"
]
},
"energy": {
"description": "Is it used to calculate job energy. Overwrite global",
"type": "string",
"enum": [
"power",
"energy"
]
},
"lowerIsBetter": {
"description": "Is lower better. Overwrite global",
"type": "boolean"
},
"peak": {
"description": "The maximum possible metric value",
"type": "number"
},
"normal": {
"description": "A common metric value level",
"type": "number"
},
"caution": {
"description": "Metric value requires attention",
"type": "number"
},
"alert": {
"description": "Metric value requiring immediate attention",
"type": "number"
},
"remove": {
"description": "Remove this metric for this subcluster",
"type": "boolean"
}
},
"required": [
"name"
]
}
}
},
"required": [
"name",
"unit",
"scope",
"timestep",
"aggregation",
"peak",
"normal",
"caution",
"alert"
]
},
"minItems": 1
},
"subClusters": {
"description": "Array of cluster hardware partitions",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "Hardware partition name",
"type": "string"
},
"processorType": {
"description": "Processor type",
"type": "string"
},
"socketsPerNode": {
"description": "Number of sockets per node",
"type": "integer"
},
"coresPerSocket": {
"description": "Number of cores per socket",
"type": "integer"
},
"threadsPerCore": {
"description": "Number of SMT threads per core",
"type": "integer"
},
"flopRateScalar": {
"description": "Theoretical node peak flop rate for scalar code in GFlops/s",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"value": {
"type": "number"
}
}
},
"flopRateSimd": {
"description": "Theoretical node peak flop rate for SIMD code in GFlops/s",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"value": {
"type": "number"
}
}
},
"memoryBandwidth": {
"description": "Theoretical node peak memory bandwidth in GB/s",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"value": {
"type": "number"
}
}
},
"nodes": {
"description": "Node list expression",
"type": "string"
},
"topology": {
"description": "Node topology",
"type": "object",
"properties": {
"node": {
"description": "HwTread lists of node",
"type": "array",
"items": {
"type": "integer"
}
},
"socket": {
"description": "HwTread lists of sockets",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"memoryDomain": {
"description": "HwTread lists of memory domains",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"die": {
"description": "HwTread lists of dies",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"core": {
"description": "HwTread lists of cores",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "integer"
}
}
},
"accelerators": {
"type": "array",
"description": "List of of accelerator devices",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "The unique device id"
},
"type": {
"type": "string",
"description": "The accelerator type",
"enum": [
"Nvidia GPU",
"AMD GPU",
"Intel GPU"
]
},
"model": {
"type": "string",
"description": "The accelerator model"
}
},
"required": [
"id",
"type",
"model"
]
}
}
},
"required": [
"node",
"socket",
"memoryDomain"
]
}
},
"required": [
"name",
"nodes",
"topology",
"processorType",
"socketsPerNode",
"coresPerSocket",
"threadsPerCore",
"flopRateScalar",
"flopRateSimd",
"memoryBandwidth"
]
},
"minItems": 1
}
},
"required": [
"name",
"metricConfig",
"subClusters"
]
}

871
pkg/schema/schemas/config.schema.json
View File

@@ -1,433 +1,498 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://config.schema.json",
"title": "cc-backend configuration file schema",
"type": "object",
"properties": {
"addr": {
"description": "Address where the http (or https) server will listen on (for example: 'localhost:80').",
"type": "string"
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://config.schema.json",
"title": "cc-backend configuration file schema",
"type": "object",
"properties": {
"addr": {
"description": "Address where the http (or https) server will listen on (for example: 'localhost:80').",
"type": "string"
},
"apiAllowedIPs": {
"description": "Addresses from which secured API endpoints can be reached",
"type": "array",
"items": {
"type": "string"
}
},
"user": {
"description": "Drop root permissions once .env was read and the port was taken. Only applicable if using privileged port.",
"type": "string"
},
"group": {
"description": "Drop root permissions once .env was read and the port was taken. Only applicable if using privileged port.",
"type": "string"
},
"disable-authentication": {
"description": "Disable authentication (for everything: API, Web-UI, ...).",
"type": "boolean"
},
"embed-static-files": {
"description": "If all files in `web/frontend/public` should be served from within the binary itself (they are embedded) or not.",
"type": "boolean"
},
"static-files": {
"description": "Folder where static assets can be found, if embed-static-files is false.",
"type": "string"
},
"db-driver": {
"description": "sqlite3 or mysql (mysql will work for mariadb as well).",
"type": "string",
"enum": [
"sqlite3",
"mysql"
]
},
"db": {
"description": "For sqlite3 a filename, for mysql a DSN in this format: https://github.com/go-sql-driver/mysql#dsn-data-source-name (Without query parameters!).",
"type": "string"
},
"archive": {
"description": "Configuration keys for job-archive",
"type": "object",
"properties": {
"kind": {
"description": "Backend type for job-archive",
"type": "string",
"enum": [
"file",
"s3"
]
},
"user": {
"description": "Drop root permissions once .env was read and the port was taken. Only applicable if using privileged port.",
"type": "string"
"path": {
"description": "Path to job archive for file backend",
"type": "string"
},
"group": {
"description": "Drop root permissions once .env was read and the port was taken. Only applicable if using privileged port.",
"type": "string"
"compression": {
"description": "Setup automatic compression for jobs older than number of days",
"type": "integer"
},
"disable-authentication": {
"description": "Disable authentication (for everything: API, Web-UI, ...).",
"type": "boolean"
},
"embed-static-files": {
"description": "If all files in `web/frontend/public` should be served from within the binary itself (they are embedded) or not.",
"type": "boolean"
},
"static-files": {
"description": "Folder where static assets can be found, if embed-static-files is false.",
"type": "string"
},
"db-driver": {
"description": "sqlite3 or mysql (mysql will work for mariadb as well).",
"type": "string",
"enum": [
"sqlite3",
"mysql"
]
},
"db": {
"description": "For sqlite3 a filename, for mysql a DSN in this format: https://github.com/go-sql-driver/mysql#dsn-data-source-name (Without query parameters!).",
"type": "string"
},
"job-archive": {
"description": "Configuration keys for job-archive",
"type": "object",
"properties": {
"kind": {
"description": "Backend type for job-archive",
"type": "string",
"enum": [
"file",
"s3"
]
},
"path": {
"description": "Path to job archive for file backend",
"type": "string"
},
"compression": {
"description": "Setup automatic compression for jobs older than number of days",
"type": "integer"
},
"retention": {
"description": "Configuration keys for retention",
"type": "object",
"properties": {
"policy": {
"description": "Retention policy",
"type": "string",
"enum": [
"none",
"delete",
"move"
]
},
"includeDB": {
"description": "Also remove jobs from database",
"type": "boolean"
},
"age": {
"description": "Act on jobs with startTime older than age (in days)",
"type": "integer"
},
"location": {
"description": "The target directory for retention. Only applicable for retention move.",
"type": "string"
}
},
"required": [
"policy"
]
}
"retention": {
"description": "Configuration keys for retention",
"type": "object",
"properties": {
"policy": {
"description": "Retention policy",
"type": "string",
"enum": [
"none",
"delete",
"move"
]
},
"required": [
"kind"
]
"includeDB": {
"description": "Also remove jobs from database",
"type": "boolean"
},
"age": {
"description": "Act on jobs with startTime older than age (in days)",
"type": "integer"
},
"location": {
"description": "The target directory for retention. Only applicable for retention move.",
"type": "string"
}
},
"required": [
"policy"
]
}
},
"required": [
"kind"
]
},
"disable-archive": {
"description": "Keep all metric data in the metric data repositories, do not write to the job-archive.",
"type": "boolean"
},
"validate": {
"description": "Validate all input json documents against json schema.",
"type": "boolean"
},
"session-max-age": {
"description": "Specifies for how long a session shall be valid as a string parsable by time.ParseDuration(). If 0 or empty, the session/token does not expire!",
"type": "string"
},
"https-cert-file": {
"description": "Filepath to SSL certificate. If also https-key-file is set use HTTPS using those certificates.",
"type": "string"
},
"https-key-file": {
"description": "Filepath to SSL key file. If also https-cert-file is set use HTTPS using those certificates.",
"type": "string"
},
"redirect-http-to": {
"description": "If not the empty string and addr does not end in :80, redirect every request incoming at port 80 to that url.",
"type": "string"
},
"stop-jobs-exceeding-walltime": {
"description": "If not zero, automatically mark jobs as stopped running X seconds longer than their walltime. Only applies if walltime is set for job.",
"type": "integer"
},
"short-running-jobs-duration": {
"description": "Do not show running jobs shorter than X seconds.",
"type": "integer"
},
"emission-constant": {
"description": ".",
"type": "integer"
},
"cron-frequency": {
"description": "Frequency of cron job workers.",
"type": "object",
"properties": {
"duration-worker": {
"description": "Duration Update Worker [Defaults to '5m']",
"type": "string"
},
"disable-archive": {
"description": "Keep all metric data in the metric data repositories, do not write to the job-archive.",
"type": "boolean"
"footprint-worker": {
"description": "Metric-Footprint Update Worker [Defaults to '10m']",
"type": "string"
}
}
},
"enable-resampling": {
"description": "Enable dynamic zoom in frontend metric plots.",
"type": "object",
"properties": {
"trigger": {
"description": "Trigger next zoom level at less than this many visible datapoints.",
"type": "integer"
},
"validate": {
"description": "Validate all input json documents against json schema.",
"type": "boolean"
},
"session-max-age": {
"description": "Specifies for how long a session shall be valid as a string parsable by time.ParseDuration(). If 0 or empty, the session/token does not expire!",
"type": "string"
},
"https-cert-file": {
"description": "Filepath to SSL certificate. If also https-key-file is set use HTTPS using those certificates.",
"type": "string"
},
"https-key-file": {
"description": "Filepath to SSL key file. If also https-cert-file is set use HTTPS using those certificates.",
"type": "string"
},
"redirect-http-to": {
"description": "If not the empty string and addr does not end in :80, redirect every request incoming at port 80 to that url.",
"type": "string"
},
"stop-jobs-exceeding-walltime": {
"description": "If not zero, automatically mark jobs as stopped running X seconds longer than their walltime. Only applies if walltime is set for job.",
"resolutions": {
"description": "Array of resampling target resolutions, in seconds.",
"type": "array",
"items": {
"type": "integer"
}
}
},
"required": [
"trigger",
"resolutions"
]
},
"jwts": {
"description": "For JWT token authentication.",
"type": "object",
"properties": {
"max-age": {
"description": "Configure how long a token is valid. As string parsable by time.ParseDuration()",
"type": "string"
},
"short-running-jobs-duration": {
"description": "Do not show running jobs shorter than X seconds.",
"type": "integer"
"cookieName": {
"description": "Cookie that should be checked for a JWT token.",
"type": "string"
},
"jwts": {
"description": "For JWT token authentication.",
"validateUser": {
"description": "Deny login for users not in database (but defined in JWT). Overwrite roles in JWT with database roles.",
"type": "boolean"
},
"trustedIssuer": {
"description": "Issuer that should be accepted when validating external JWTs ",
"type": "string"
},
"syncUserOnLogin": {
"description": "Add non-existent user to DB at login attempt with values provided in JWT.",
"type": "boolean"
}
},
"required": [
"max-age"
]
},
"oidc": {
"provider": {
"description": "",
"type": "string"
},
"syncUserOnLogin": {
"description": "",
"type": "boolean"
},
"updateUserOnLogin": {
"description": "",
"type": "boolean"
},
"required": [
"provider"
]
},
"ldap": {
"description": "For LDAP Authentication and user synchronisation.",
"type": "object",
"properties": {
"url": {
"description": "URL of LDAP directory server.",
"type": "string"
},
"user_base": {
"description": "Base DN of user tree root.",
"type": "string"
},
"search_dn": {
"description": "DN for authenticating LDAP admin account with general read rights.",
"type": "string"
},
"user_bind": {
"description": "Expression used to authenticate users via LDAP bind. Must contain uid={username}.",
"type": "string"
},
"user_filter": {
"description": "Filter to extract users for syncing.",
"type": "string"
},
"username_attr": {
"description": "Attribute with full username. Default: gecos",
"type": "string"
},
"sync_interval": {
"description": "Interval used for syncing local user table with LDAP directory. Parsed using time.ParseDuration.",
"type": "string"
},
"sync_del_old_users": {
"description": "Delete obsolete users in database.",
"type": "boolean"
},
"syncUserOnLogin": {
"description": "Add non-existent user to DB at login attempt if user exists in Ldap directory",
"type": "boolean"
}
},
"required": [
"url",
"user_base",
"search_dn",
"user_bind",
"user_filter"
]
},
"clusters": {
"description": "Configuration for the clusters to be displayed.",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"description": "The name of the cluster.",
"type": "string"
},
"metricDataRepository": {
"description": "Type of the metric data repository for this cluster",
"type": "object",
"properties": {
"max-age": {
"description": "Configure how long a token is valid. As string parsable by time.ParseDuration()",
"type": "string"
},
"cookieName": {
"description": "Cookie that should be checked for a JWT token.",
"type": "string"
},
"validateUser": {
"description": "Deny login for users not in database (but defined in JWT). Overwrite roles in JWT with database roles.",
"type": "boolean"
},
"trustedIssuer": {
"description": "Issuer that should be accepted when validating external JWTs ",
"type": "string"
},
"syncUserOnLogin": {
"description": "Add non-existent user to DB at login attempt with values provided in JWT.",
"type": "boolean"
}
"kind": {
"type": "string",
"enum": [
"influxdb",
"prometheus",
"cc-metric-store",
"test"
]
},
"url": {
"type": "string"
},
"token": {
"type": "string"
}
},
"required": [
"max-age"
"kind",
"url"
]
},
"ldap": {
"description": "For LDAP Authentication and user synchronisation.",
},
"filterRanges": {
"description": "This option controls the slider ranges for the UI controls of numNodes, duration, and startTime.",
"type": "object",
"properties": {
"url": {
"description": "URL of LDAP directory server.",
"type": "string"
},
"user_base": {
"description": "Base DN of user tree root.",
"type": "string"
},
"search_dn": {
"description": "DN for authenticating LDAP admin account with general read rights.",
"type": "string"
},
"user_bind": {
"description": "Expression used to authenticate users via LDAP bind. Must contain uid={username}.",
"type": "string"
},
"user_filter": {
"description": "Filter to extract users for syncing.",
"type": "string"
},
"username_attr": {
"description": "Attribute with full username. Default: gecos",
"type": "string"
},
"sync_interval": {
"description": "Interval used for syncing local user table with LDAP directory. Parsed using time.ParseDuration.",
"type": "string"
},
"sync_del_old_users": {
"description": "Delete obsolete users in database.",
"type": "boolean"
},
"syncUserOnLogin": {
"description": "Add non-existent user to DB at login attempt if user exists in Ldap directory",
"type": "boolean"
}
},
"required": [
"url",
"user_base",
"search_dn",
"user_bind",
"user_filter"
]
},
"clusters": {
"description": "Configuration for the clusters to be displayed.",
"type": "array",
"items": {
"numNodes": {
"description": "UI slider range for number of nodes",
"type": "object",
"properties": {
"name": {
"description": "The name of the cluster.",
"type": "string"
},
"metricDataRepository": {
"description": "Type of the metric data repository for this cluster",
"type": "object",
"properties": {
"kind": {
"type": "string",
"enum": [
"influxdb",
"prometheus",
"cc-metric-store",
"test"
]
},
"url": {
"type": "string"
},
"token": {
"type": "string"
}
},
"required": [
"kind",
"url"
]
},
"filterRanges": {
"description": "This option controls the slider ranges for the UI controls of numNodes, duration, and startTime.",
"type": "object",
"properties": {
"numNodes": {
"description": "UI slider range for number of nodes",
"type": "object",
"properties": {
"from": {
"type": "integer"
},
"to": {
"type": "integer"
}
},
"required": [
"from",
"to"
]
},
"duration": {
"description": "UI slider range for duration",
"type": "object",
"properties": {
"from": {
"type": "integer"
},
"to": {
"type": "integer"
}
},
"required": [
"from",
"to"
]
},
"startTime": {
"description": "UI slider range for start time",
"type": "object",
"properties": {
"from": {
"type": "string",
"format": "date-time"
},
"to": {
"type": "null"
}
},
"required": [
"from",
"to"
]
}
},
"required": [
"numNodes",
"duration",
"startTime"
]
}
"from": {
"type": "integer"
},
"to": {
"type": "integer"
}
},
"required": [
"name",
"metricDataRepository",
"filterRanges"
],
"minItems": 1
}
},
"ui-defaults": {
"description": "Default configuration for web UI",
"type": "object",
"properties": {
"plot_general_colorBackground": {
"description": "Color plot background according to job average threshold limits",
"type": "boolean"
},
"plot_general_lineWidth": {
"description": "Initial linewidth",
"from",
"to"
]
},
"duration": {
"description": "UI slider range for duration",
"type": "object",
"properties": {
"from": {
"type": "integer"
},
"plot_list_jobsPerPage": {
"description": "Jobs shown per page in job lists",
},
"to": {
"type": "integer"
}
},
"plot_view_plotsPerRow": {
"description": "Number of plots per row in single job view",
"type": "integer"
"required": [
"from",
"to"
]
},
"startTime": {
"description": "UI slider range for start time",
"type": "object",
"properties": {
"from": {
"type": "string",
"format": "date-time"
},
"to": {
"type": "null"
}
},
"plot_view_showPolarplot": {
"description": "Option to toggle polar plot in single job view",
"type": "boolean"
},
"plot_view_showRoofline": {
"description": "Option to toggle roofline plot in single job view",
"type": "boolean"
},
"plot_view_showStatTable": {
"description": "Option to toggle the node statistic table in single job view",
"type": "boolean"
},
"system_view_selectedMetric": {
"description": "Initial metric shown in system view",
"type": "string"
},
"analysis_view_histogramMetrics": {
"description": "Metrics to show as job count histograms in analysis view",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"analysis_view_scatterPlotMetrics": {
"description": "Initial scatter plto configuration in analysis view",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "string",
"minItems": 2,
"maxItems": 2
},
"minItems": 1
}
},
"job_view_nodestats_selectedMetrics": {
"description": "Initial metrics shown in node statistics table of single job view",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"job_view_polarPlotMetrics": {
"description": "Metrics shown in polar plot of single job view",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"job_view_selectedMetrics": {
"description": "",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"plot_general_colorscheme": {
"description": "Initial color scheme",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"plot_list_selectedMetrics": {
"description": "Initial metric plots shown in jobs lists",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
}
"required": [
"from",
"to"
]
}
},
"required": [
"plot_general_colorBackground",
"plot_general_lineWidth",
"plot_list_jobsPerPage",
"plot_view_plotsPerRow",
"plot_view_showPolarplot",
"plot_view_showRoofline",
"plot_view_showStatTable",
"system_view_selectedMetric",
"analysis_view_histogramMetrics",
"analysis_view_scatterPlotMetrics",
"job_view_nodestats_selectedMetrics",
"job_view_polarPlotMetrics",
"job_view_selectedMetrics",
"plot_general_colorscheme",
"plot_list_selectedMetrics"
"numNodes",
"duration",
"startTime"
]
}
}
},
"required": [
"name",
"metricDataRepository",
"filterRanges"
],
"minItems": 1
}
},
"required": [
"jwts",
"clusters"
]
"ui-defaults": {
"description": "Default configuration for web UI",
"type": "object",
"properties": {
"plot_general_colorBackground": {
"description": "Color plot background according to job average threshold limits",
"type": "boolean"
},
"plot_general_lineWidth": {
"description": "Initial linewidth",
"type": "integer"
},
"plot_list_jobsPerPage": {
"description": "Jobs shown per page in job lists",
"type": "integer"
},
"plot_view_plotsPerRow": {
"description": "Number of plots per row in single job view",
"type": "integer"
},
"plot_view_showPolarplot": {
"description": "Option to toggle polar plot in single job view",
"type": "boolean"
},
"plot_view_showRoofline": {
"description": "Option to toggle roofline plot in single job view",
"type": "boolean"
},
"plot_view_showStatTable": {
"description": "Option to toggle the node statistic table in single job view",
"type": "boolean"
},
"system_view_selectedMetric": {
"description": "Initial metric shown in system view",
"type": "string"
},
"job_view_showFootprint": {
"description": "Option to toggle footprint ui in single job view",
"type": "boolean"
},
"job_list_usePaging": {
"description": "Option to switch from continous scroll to paging",
"type": "boolean"
},
"analysis_view_histogramMetrics": {
"description": "Metrics to show as job count histograms in analysis view",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"analysis_view_scatterPlotMetrics": {
"description": "Initial scatter plto configuration in analysis view",
"type": "array",
"items": {
"type": "array",
"items": {
"type": "string",
"minItems": 2,
"maxItems": 2
},
"minItems": 1
}
},
"job_view_nodestats_selectedMetrics": {
"description": "Initial metrics shown in node statistics table of single job view",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"job_view_selectedMetrics": {
"description": "Initial metrics shown as plots in single job view",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"plot_general_colorscheme": {
"description": "Initial color scheme",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
},
"plot_list_selectedMetrics": {
"description": "Initial metric plots shown in jobs lists",
"type": "array",
"items": {
"type": "string",
"minItems": 1
}
}
},
"required": [
"plot_general_colorBackground",
"plot_general_lineWidth",
"plot_list_jobsPerPage",
"plot_view_plotsPerRow",
"plot_view_showPolarplot",
"plot_view_showRoofline",
"plot_view_showStatTable",
"system_view_selectedMetric",
"job_view_showFootprint",
"job_list_usePaging",
"analysis_view_histogramMetrics",
"analysis_view_scatterPlotMetrics",
"job_view_nodestats_selectedMetrics",
"job_view_selectedMetrics",
"plot_general_colorscheme",
"plot_list_selectedMetrics"
]
}
},
"required": [
"jwts",
"clusters",
"apiAllowedIPs"
]
}

956
pkg/schema/schemas/job-data.schema.json
View File

@@ -1,490 +1,490 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-data.schema.json",
"title": "Job metric data list",
"description": "Collection of metric data of a HPC job",
"type": "object",
"properties": {
"mem_used": {
"description": "Memory capacity used",
"type": "object",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"flops_any": {
"description": "Total flop rate with DP flops scaled up",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"mem_bw": {
"description": "Main memory bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"net_bw": {
"description": "Total fast interconnect network bandwidth",
"type": "object",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ipc": {
"description": "Instructions executed per cycle",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"cpu_user": {
"description": "CPU user active core utilization",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"cpu_load": {
"description": "CPU requested core utilization (load 1m)",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"flops_dp": {
"description": "Double precision flop rate",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"flops_sp": {
"description": "Single precision flops rate",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"vectorization_ratio": {
"description": "Fraction of arithmetic instructions using SIMD instructions",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"cpu_power": {
"description": "CPU power consumption",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"mem_power": {
"description": "Memory power consumption",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"acc_utilization": {
"description": "GPU utilization",
"properties": {
"accelerator": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"accelerator"
]
},
"acc_mem_used": {
"description": "GPU memory capacity used",
"properties": {
"accelerator": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"accelerator"
]
},
"acc_power": {
"description": "GPU power consumption",
"properties": {
"accelerator": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"accelerator"
]
},
"clock": {
"description": "Average core frequency",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"eth_read_bw": {
"description": "Ethernet read bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"eth_write_bw": {
"description": "Ethernet write bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"filesystems": {
"description": "Array of filesystems",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"type": {
"type": "string",
"enum": [
"nfs",
"lustre",
"gpfs",
"nvme",
"ssd",
"hdd",
"beegfs"
]
},
"read_bw": {
"description": "File system read bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"write_bw": {
"description": "File system write bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"read_req": {
"description": "File system read requests",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"write_req": {
"description": "File system write requests",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"inodes": {
"description": "File system write requests",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"accesses": {
"description": "File system open and close",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"fsync": {
"description": "File system fsync",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"create": {
"description": "File system create",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"open": {
"description": "File system open",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"close": {
"description": "File system close",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"seek": {
"description": "File system seek",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
}
},
"required": [
"name",
"type",
"read_bw",
"write_bw"
]
},
"minItems": 1
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-data.schema.json",
"title": "Job metric data list",
"description": "Collection of metric data of a HPC job",
"type": "object",
"properties": {
"mem_used": {
"description": "Memory capacity used",
"type": "object",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ic_rcv_packets": {
"description": "Network interconnect read packets",
"flops_any": {
"description": "Total flop rate with DP flops scaled up",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"mem_bw": {
"description": "Main memory bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"net_bw": {
"description": "Total fast interconnect network bandwidth",
"type": "object",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ipc": {
"description": "Instructions executed per cycle",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"cpu_user": {
"description": "CPU user active core utilization",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"cpu_load": {
"description": "CPU requested core utilization (load 1m)",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"flops_dp": {
"description": "Double precision flop rate",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"flops_sp": {
"description": "Single precision flops rate",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"vectorization_ratio": {
"description": "Fraction of arithmetic instructions using SIMD instructions",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"cpu_power": {
"description": "CPU power consumption",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"mem_power": {
"description": "Memory power consumption",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"acc_utilization": {
"description": "GPU utilization",
"properties": {
"accelerator": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"accelerator"
]
},
"acc_mem_used": {
"description": "GPU memory capacity used",
"properties": {
"accelerator": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"accelerator"
]
},
"acc_power": {
"description": "GPU power consumption",
"properties": {
"accelerator": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"accelerator"
]
},
"clock": {
"description": "Average core frequency",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"socket": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"memoryDomain": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"core": {
"$ref": "embedfs://job-metric-data.schema.json"
},
"hwthread": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"minProperties": 1
},
"eth_read_bw": {
"description": "Ethernet read bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"eth_write_bw": {
"description": "Ethernet write bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"filesystems": {
"description": "Array of filesystems",
"type": "array",
"items": {
"type": "object",
"properties": {
"node": {
"name": {
"type": "string"
},
"type": {
"type": "string",
"enum": [
"nfs",
"lustre",
"gpfs",
"nvme",
"ssd",
"hdd",
"beegfs"
]
},
"read_bw": {
"description": "File system read bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
}
},
"required": [
"node"
]
},
"write_bw": {
"description": "File system write bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"read_req": {
"description": "File system read requests",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"write_req": {
"description": "File system write requests",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"inodes": {
"description": "File system write requests",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"accesses": {
"description": "File system open and close",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"fsync": {
"description": "File system fsync",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"create": {
"description": "File system create",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"open": {
"description": "File system open",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"close": {
"description": "File system close",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"seek": {
"description": "File system seek",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
}
},
"required": [
"node"
]
},
"ic_send_packets": {
"description": "Network interconnect send packet",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ic_read_bw": {
"description": "Network interconnect read bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ic_write_bw": {
"description": "Network interconnect write bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
"name",
"type",
"read_bw",
"write_bw"
]
},
"minItems": 1
}
},
"ic_rcv_packets": {
"description": "Network interconnect read packets",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"cpu_user",
"cpu_load",
"mem_used",
"flops_any",
"mem_bw",
"net_bw",
"filesystems"
"node"
]
},
"ic_send_packets": {
"description": "Network interconnect send packet",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ic_read_bw": {
"description": "Network interconnect read bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"ic_write_bw": {
"description": "Network interconnect write bandwidth",
"properties": {
"node": {
"$ref": "embedfs://job-metric-data.schema.json"
}
},
"required": [
"node"
]
},
"required": [
"cpu_user",
"cpu_load",
"mem_used",
"flops_any",
"mem_bw",
"net_bw",
"filesystems"
]
}

672
pkg/schema/schemas/job-meta.schema.json
View File

@@ -1,351 +1,351 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-meta.schema.json",
"title": "Job meta data",
"description": "Meta data information of a HPC job",
"type": "object",
"properties": {
"jobId": {
"description": "The unique identifier of a job",
"type": "integer"
},
"user": {
"description": "The unique identifier of a user",
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-meta.schema.json",
"title": "Job meta data",
"description": "Meta data information of a HPC job",
"type": "object",
"properties": {
"jobId": {
"description": "The unique identifier of a job",
"type": "integer"
},
"user": {
"description": "The unique identifier of a user",
"type": "string"
},
"project": {
"description": "The unique identifier of a project",
"type": "string"
},
"cluster": {
"description": "The unique identifier of a cluster",
"type": "string"
},
"subCluster": {
"description": "The unique identifier of a sub cluster",
"type": "string"
},
"partition": {
"description": "The Slurm partition to which the job was submitted",
"type": "string"
},
"arrayJobId": {
"description": "The unique identifier of an array job",
"type": "integer"
},
"numNodes": {
"description": "Number of nodes used",
"type": "integer",
"exclusiveMinimum": 0
},
"numHwthreads": {
"description": "Number of HWThreads used",
"type": "integer",
"exclusiveMinimum": 0
},
"numAcc": {
"description": "Number of accelerators used",
"type": "integer",
"exclusiveMinimum": 0
},
"exclusive": {
"description": "Specifies how nodes are shared. 0 - Shared among multiple jobs of multiple users, 1 - Job exclusive, 2 - Shared among multiple jobs of same user",
"type": "integer",
"minimum": 0,
"maximum": 2
},
"monitoringStatus": {
"description": "State of monitoring system during job run",
"type": "integer"
},
"smt": {
"description": "SMT threads used by job",
"type": "integer"
},
"walltime": {
"description": "Requested walltime of job in seconds",
"type": "integer",
"exclusiveMinimum": 0
},
"jobState": {
"description": "Final state of job",
"type": "string",
"enum": [
"completed",
"failed",
"cancelled",
"stopped",
"out_of_memory",
"timeout"
]
},
"startTime": {
"description": "Start epoch time stamp in seconds",
"type": "integer",
"exclusiveMinimum": 0
},
"duration": {
"description": "Duration of job in seconds",
"type": "integer",
"exclusiveMinimum": 0
},
"resources": {
"description": "Resources used by job",
"type": "array",
"items": {
"type": "object",
"properties": {
"hostname": {
"type": "string"
},
"project": {
"description": "The unique identifier of a project",
"type": "string"
},
"cluster": {
"description": "The unique identifier of a cluster",
"type": "string"
},
"subCluster": {
"description": "The unique identifier of a sub cluster",
"type": "string"
},
"partition": {
"description": "The Slurm partition to which the job was submitted",
"type": "string"
},
"arrayJobId": {
"description": "The unique identifier of an array job",
"type": "integer"
},
"numNodes": {
"description": "Number of nodes used",
"type": "integer",
"exclusiveMinimum": 0
},
"numHwthreads": {
"description": "Number of HWThreads used",
"type": "integer",
"exclusiveMinimum": 0
},
"numAcc": {
"description": "Number of accelerators used",
"type": "integer",
"exclusiveMinimum": 0
},
"exclusive": {
"description": "Specifies how nodes are shared. 0 - Shared among multiple jobs of multiple users, 1 - Job exclusive, 2 - Shared among multiple jobs of same user",
"type": "integer",
"minimum": 0,
"maximum": 2
},
"monitoringStatus": {
"description": "State of monitoring system during job run",
"type": "integer"
},
"smt": {
"description": "SMT threads used by job",
"type": "integer"
},
"walltime": {
"description": "Requested walltime of job in seconds",
"type": "integer",
"exclusiveMinimum": 0
},
"jobState": {
"description": "Final state of job",
},
"hwthreads": {
"type": "array",
"description": "List of OS processor ids",
"items": {
"type": "integer"
}
},
"accelerators": {
"type": "array",
"description": "List of of accelerator device ids",
"items": {
"type": "string"
}
},
"configuration": {
"type": "string",
"enum": [
"completed",
"failed",
"cancelled",
"stopped",
"out_of_memory",
"timeout"
]
"description": "The configuration options of the node"
}
},
"startTime": {
"description": "Start epoch time stamp in seconds",
"type": "integer",
"exclusiveMinimum": 0
"required": [
"hostname"
],
"minItems": 1
}
},
"metaData": {
"description": "Additional information about the job",
"type": "object",
"properties": {
"jobScript": {
"type": "string",
"description": "The batch script of the job"
},
"duration": {
"description": "Duration of job in seconds",
"type": "integer",
"exclusiveMinimum": 0
"jobName": {
"type": "string",
"description": "Slurm Job name"
},
"resources": {
"description": "Resources used by job",
"type": "array",
"items": {
"type": "object",
"properties": {
"hostname": {
"type": "string"
},
"hwthreads": {
"type": "array",
"description": "List of OS processor ids",
"items": {
"type": "integer"
}
},
"accelerators": {
"type": "array",
"description": "List of of accelerator device ids",
"items": {
"type": "string"
}
},
"configuration": {
"type": "string",
"description": "The configuration options of the node"
}
},
"required": [
"hostname"
],
"minItems": 1
}
"slurmInfo": {
"type": "string",
"description": "Additional slurm infos as show by scontrol show job"
}
}
},
"tags": {
"description": "List of tags",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"type": {
"type": "string"
}
},
"metaData": {
"description": "Additional information about the job",
"required": [
"name",
"type"
]
},
"uniqueItems": true
},
"statistics": {
"description": "Job statistic data",
"type": "object",
"properties": {
"mem_used": {
"description": "Memory capacity used (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"cpu_load": {
"description": "CPU requested core utilization (load 1m) (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"flops_any": {
"description": "Total flop rate with DP flops scaled up (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"mem_bw": {
"description": "Main memory bandwidth (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"net_bw": {
"description": "Total fast interconnect network bandwidth (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"file_bw": {
"description": "Total file IO bandwidth (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ipc": {
"description": "Instructions executed per cycle",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"cpu_user": {
"description": "CPU user active core utilization",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"flops_dp": {
"description": "Double precision flop rate",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"flops_sp": {
"description": "Single precision flops rate",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"rapl_power": {
"description": "CPU power consumption",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"acc_used": {
"description": "GPU utilization",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"acc_mem_used": {
"description": "GPU memory capacity used",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"acc_power": {
"description": "GPU power consumption",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"clock": {
"description": "Average core frequency",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"eth_read_bw": {
"description": "Ethernet read bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"eth_write_bw": {
"description": "Ethernet write bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_rcv_packets": {
"description": "Network interconnect read packets",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_send_packets": {
"description": "Network interconnect send packet",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_read_bw": {
"description": "Network interconnect read bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_write_bw": {
"description": "Network interconnect write bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"filesystems": {
"description": "Array of filesystems",
"type": "array",
"items": {
"type": "object",
"properties": {
"jobScript": {
"type": "string",
"description": "The batch script of the job"
},
"jobName": {
"type": "string",
"description": "Slurm Job name"
},
"slurmInfo": {
"type": "string",
"description": "Additional slurm infos as show by scontrol show job"
}
}
},
"tags": {
"description": "List of tags",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"type": {
"type": "string"
}
},
"required": [
"name",
"type"
"name": {
"type": "string"
},
"type": {
"type": "string",
"enum": [
"nfs",
"lustre",
"gpfs",
"nvme",
"ssd",
"hdd",
"beegfs"
]
},
"uniqueItems": true
},
"statistics": {
"description": "Job statistic data",
"type": "object",
"properties": {
"mem_used": {
"description": "Memory capacity used (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"cpu_load": {
"description": "CPU requested core utilization (load 1m) (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"flops_any": {
"description": "Total flop rate with DP flops scaled up (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"mem_bw": {
"description": "Main memory bandwidth (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"net_bw": {
"description": "Total fast interconnect network bandwidth (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"file_bw": {
"description": "Total file IO bandwidth (required)",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ipc": {
"description": "Instructions executed per cycle",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"cpu_user": {
"description": "CPU user active core utilization",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"flops_dp": {
"description": "Double precision flop rate",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"flops_sp": {
"description": "Single precision flops rate",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"rapl_power": {
"description": "CPU power consumption",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"acc_used": {
"description": "GPU utilization",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"acc_mem_used": {
"description": "GPU memory capacity used",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"acc_power": {
"description": "GPU power consumption",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"clock": {
"description": "Average core frequency",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"eth_read_bw": {
"description": "Ethernet read bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"eth_write_bw": {
"description": "Ethernet write bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_rcv_packets": {
"description": "Network interconnect read packets",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_send_packets": {
"description": "Network interconnect send packet",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_read_bw": {
"description": "Network interconnect read bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"ic_write_bw": {
"description": "Network interconnect write bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"filesystems": {
"description": "Array of filesystems",
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"type": {
"type": "string",
"enum": [
"nfs",
"lustre",
"gpfs",
"nvme",
"ssd",
"hdd",
"beegfs"
]
},
"read_bw": {
"description": "File system read bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"write_bw": {
"description": "File system write bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"read_req": {
"description": "File system read requests",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"write_req": {
"description": "File system write requests",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"inodes": {
"description": "File system write requests",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"accesses": {
"description": "File system open and close",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"fsync": {
"description": "File system fsync",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"create": {
"description": "File system create",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"open": {
"description": "File system open",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"close": {
"description": "File system close",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"seek": {
"description": "File system seek",
"$ref": "embedfs://job-metric-statistics.schema.json"
}
},
"required": [
"name",
"type",
"read_bw",
"write_bw"
]
},
"minItems": 1
}
},
"read_bw": {
"description": "File system read bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"write_bw": {
"description": "File system write bandwidth",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"read_req": {
"description": "File system read requests",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"write_req": {
"description": "File system write requests",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"inodes": {
"description": "File system write requests",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"accesses": {
"description": "File system open and close",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"fsync": {
"description": "File system fsync",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"create": {
"description": "File system create",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"open": {
"description": "File system open",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"close": {
"description": "File system close",
"$ref": "embedfs://job-metric-statistics.schema.json"
},
"seek": {
"description": "File system seek",
"$ref": "embedfs://job-metric-statistics.schema.json"
}
},
"required": [
"cpu_user",
"cpu_load",
"mem_used",
"flops_any",
"mem_bw"
"name",
"type",
"read_bw",
"write_bw"
]
},
"minItems": 1
}
},
"required": [
"jobId",
"user",
"project",
"cluster",
"subCluster",
"numNodes",
"exclusive",
"startTime",
"jobState",
"duration",
"resources",
"statistics"
]
},
"required": [
"cpu_user",
"cpu_load",
"mem_used",
"flops_any",
"mem_bw"
]
}
},
"required": [
"jobId",
"user",
"project",
"cluster",
"subCluster",
"numNodes",
"exclusive",
"startTime",
"jobState",
"duration",
"resources",
"statistics"
]
}

426
pkg/schema/schemas/job-metric-data.schema.json
View File

@@ -1,216 +1,216 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-metric-data.schema.json",
"title": "Job metric data",
"description": "Metric data of a HPC job",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"timestep": {
"description": "Measurement interval in seconds",
"type": "integer"
},
"thresholds": {
"description": "Metric thresholds for specific system",
"type": "object",
"properties": {
"peak": {
"type": "number"
},
"normal": {
"type": "number"
},
"caution": {
"type": "number"
},
"alert": {
"type": "number"
}
}
},
"statisticsSeries": {
"type": "object",
"description": "Statistics series across topology",
"properties": {
"min": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"max": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"mean": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"percentiles": {
"type": "object",
"properties": {
"10": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"20": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"30": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"40": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"50": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"60": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"70": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"80": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"90": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"25": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"75": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
}
}
}
}
},
"series": {
"type": "array",
"items": {
"type": "object",
"properties": {
"hostname": {
"type": "string"
},
"id": {
"type": "string"
},
"statistics": {
"type": "object",
"description": "Statistics across time dimension",
"properties": {
"avg": {
"description": "Series average",
"type": "number",
"minimum": 0
},
"min": {
"description": "Series minimum",
"type": "number",
"minimum": 0
},
"max": {
"description": "Series maximum",
"type": "number",
"minimum": 0
}
},
"required": [
"avg",
"min",
"max"
]
},
"data": {
"type": "array",
"contains": {
"type": "number",
"minimum": 0
},
"minItems": 1
}
},
"required": [
"hostname",
"statistics",
"data"
]
}
}
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-metric-data.schema.json",
"title": "Job metric data",
"description": "Metric data of a HPC job",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"required": [
"unit",
"timestep",
"series"
]
"timestep": {
"description": "Measurement interval in seconds",
"type": "integer"
},
"thresholds": {
"description": "Metric thresholds for specific system",
"type": "object",
"properties": {
"peak": {
"type": "number"
},
"normal": {
"type": "number"
},
"caution": {
"type": "number"
},
"alert": {
"type": "number"
}
}
},
"statisticsSeries": {
"type": "object",
"description": "Statistics series across topology",
"properties": {
"min": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"max": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"mean": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"percentiles": {
"type": "object",
"properties": {
"10": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"20": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"30": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"40": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"50": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"60": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"70": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"80": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"90": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"25": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
},
"75": {
"type": "array",
"items": {
"type": "number",
"minimum": 0
},
"minItems": 3
}
}
}
}
},
"series": {
"type": "array",
"items": {
"type": "object",
"properties": {
"hostname": {
"type": "string"
},
"id": {
"type": "string"
},
"statistics": {
"type": "object",
"description": "Statistics across time dimension",
"properties": {
"avg": {
"description": "Series average",
"type": "number",
"minimum": 0
},
"min": {
"description": "Series minimum",
"type": "number",
"minimum": 0
},
"max": {
"description": "Series maximum",
"type": "number",
"minimum": 0
}
},
"required": [
"avg",
"min",
"max"
]
},
"data": {
"type": "array",
"contains": {
"type": "number",
"minimum": 0
},
"minItems": 1
}
},
"required": [
"hostname",
"statistics",
"data"
]
}
}
},
"required": [
"unit",
"timestep",
"series"
]
}

62
pkg/schema/schemas/job-metric-statistics.schema.json
View File

@@ -1,34 +1,34 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-metric-statistics.schema.json",
"title": "Job statistics",
"description": "Format specification for job metric statistics",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"avg": {
"description": "Job metric average",
"type": "number",
"minimum": 0
},
"min": {
"description": "Job metric minimum",
"type": "number",
"minimum": 0
},
"max": {
"description": "Job metric maximum",
"type": "number",
"minimum": 0
}
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://job-metric-statistics.schema.json",
"title": "Job statistics",
"description": "Format specification for job metric statistics",
"type": "object",
"properties": {
"unit": {
"description": "Metric unit",
"$ref": "embedfs://unit.schema.json"
},
"required": [
"unit",
"avg",
"min",
"max"
]
"avg": {
"description": "Job metric average",
"type": "number",
"minimum": 0
},
"min": {
"description": "Job metric minimum",
"type": "number",
"minimum": 0
},
"max": {
"description": "Job metric maximum",
"type": "number",
"minimum": 0
}
},
"required": [
"unit",
"avg",
"min",
"max"
]
}

75
pkg/schema/schemas/unit.schema.json
View File

@@ -1,40 +1,41 @@
{
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://unit.schema.json",
"title": "Metric unit",
"description": "Format specification for job metric units",
"type": "object",
"properties": {
"base": {
"description": "Metric base unit",
"type": "string",
"enum": [
"B",
"F",
"B/s",
"F/s",
"CPI",
"IPC",
"Hz",
"W",
"°C",
""
]
},
"prefix": {
"description": "Unit prefix",
"type": "string",
"enum": [
"K",
"M",
"G",
"T",
"P",
"E"
]
}
"$schema": "http://json-schema.org/draft/2020-12/schema",
"$id": "embedfs://unit.schema.json",
"title": "Metric unit",
"description": "Format specification for job metric units",
"type": "object",
"properties": {
"base": {
"description": "Metric base unit",
"type": "string",
"enum": [
"B",
"F",
"B/s",
"F/s",
"CPI",
"IPC",
"Hz",
"W",
"J",
"°C",
""
]
},
"required": [
"base"
]
"prefix": {
"description": "Unit prefix",
"type": "string",
"enum": [
"K",
"M",
"G",
"T",
"P",
"E"
]
}
},
"required": [
"base"
]
}

15
pkg/schema/user.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2023 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -27,6 +27,7 @@ const (
AuthViaLocalPassword AuthSource = iota
AuthViaLDAP
AuthViaToken
AuthViaOIDC
AuthViaAll
)
@@ -41,11 +42,11 @@ type User struct {
Username string `json:"username"`
Password string `json:"-"`
Name string `json:"name"`
Email string `json:"email"`
Roles []string `json:"roles"`
Projects []string `json:"projects"`
AuthType AuthType `json:"authType"`
AuthSource AuthSource `json:"authSource"`
Email string `json:"email"`
Projects []string `json:"projects"`
}
func (u *User) HasProject(project string) bool {
@@ -84,6 +85,7 @@ func IsValidRole(role string) bool {
return getRoleEnum(role) != RoleError
}
// Check if User has SPECIFIED role AND role is VALID
func (u *User) HasValidRole(role string) (hasRole bool, isValid bool) {
if IsValidRole(role) {
for _, r := range u.Roles {
@@ -96,6 +98,7 @@ func (u *User) HasValidRole(role string) (hasRole bool, isValid bool) {
return false, false
}
// Check if User has SPECIFIED role
func (u *User) HasRole(role Role) bool {
for _, r := range u.Roles {
if r == GetRoleString(role) {
@@ -105,7 +108,7 @@ func (u *User) HasRole(role Role) bool {
return false
}
// Role-Arrays are short: performance not impacted by nested loop
// Check if User has ANY of the listed roles
func (u *User) HasAnyRole(queryroles []Role) bool {
for _, ur := range u.Roles {
for _, qr := range queryroles {
@@ -117,7 +120,7 @@ func (u *User) HasAnyRole(queryroles []Role) bool {
return false
}
// Role-Arrays are short: performance not impacted by nested loop
// Check if User has ALL of the listed roles
func (u *User) HasAllRoles(queryroles []Role) bool {
target := len(queryroles)
matches := 0
@@ -137,7 +140,7 @@ func (u *User) HasAllRoles(queryroles []Role) bool {
}
}
// Role-Arrays are short: performance not impacted by nested loop
// Check if User has NONE of the listed roles
func (u *User) HasNotRoles(queryroles []Role) bool {
matches := 0
for _, ur := range u.Roles {

2
pkg/schema/user_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

2
pkg/schema/validate.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.

28
pkg/schema/validate_test.go
View File

@@ -1,4 +1,4 @@
// Copyright (C) 2022 NHR@FAU, University Erlangen-Nuremberg.
// Copyright (C) NHR@FAU, University Erlangen-Nuremberg.
// All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
@@ -14,17 +14,20 @@ func TestValidateConfig(t *testing.T) {
"jwts": {
"max-age": "2m"
},
"clusters": [
{
"name": "testcluster",
"metricDataRepository": {
"kind": "cc-metric-store",
"url": "localhost:8082"},
"filterRanges": {
"numNodes": { "from": 1, "to": 64 },
"duration": { "from": 0, "to": 86400 },
"startTime": { "from": "2022-01-01T00:00:00Z", "to": null }
}}]
"apiAllowedIPs": [
"*"
],
"clusters": [
{
"name": "testcluster",
"metricDataRepository": {
"kind": "cc-metric-store",
"url": "localhost:8082"},
"filterRanges": {
"numNodes": { "from": 1, "to": 64 },
"duration": { "from": 0, "to": 86400 },
"startTime": { "from": "2022-01-01T00:00:00Z", "to": null }
}}]
}`)
if err := Validate(Config, bytes.NewReader(json)); err != nil {
@@ -33,7 +36,6 @@ func TestValidateConfig(t *testing.T) {
}
func TestValidateJobMeta(t *testing.T) {
}
func TestValidateCluster(t *testing.T) {