continue working on non-node scoped metrics

This commit is contained in:
Lou Knauer 2022-01-10 16:13:40 +01:00
parent 1c6ab3d062
commit b7432fca5f
5 changed files with 56 additions and 24 deletions

View File

@ -88,7 +88,15 @@ func initDB(db *sqlx.DB, archive string) error {
return err
}
stmt, err := tx.PrepareNamed(schema.JobInsertStmt)
stmt, err := tx.PrepareNamed(`INSERT INTO job (
job_id, user, project, cluster, partition, array_job_id, num_nodes, num_hwthreads, num_acc,
exclusive, monitoring_status, smt, job_state, start_time, duration, resources, meta_data,
mem_used_max, flops_any_avg, mem_bw_avg, load_avg, net_bw_avg, net_data_vol_total, file_bw_avg, file_data_vol_total
) VALUES (
:job_id, :user, :project, :cluster, :partition, :array_job_id, :num_nodes, :num_hwthreads, :num_acc,
:exclusive, :monitoring_status, :smt, :job_state, :start_time, :duration, :resources, :meta_data,
:mem_used_max, :flops_any_avg, :mem_bw_avg, :load_avg, :net_bw_avg, :net_data_vol_total, :file_bw_avg, :file_data_vol_total
);`)
if err != nil {
return err
}

View File

@ -243,7 +243,7 @@ func calcStatisticsSeries(job *schema.Job, jobData schema.JobData) error {
for i := 0; i < n; i++ {
sum, smin, smax := schema.Float(0.), math.MaxFloat32, -math.MaxFloat32
for _, series := range jobMetric.Series {
if len(series.Data) >= i {
if i >= len(series.Data) {
sum, smin, smax = schema.NaN, math.NaN(), math.NaN()
break
}
@ -258,9 +258,9 @@ func calcStatisticsSeries(job *schema.Job, jobData schema.JobData) error {
max[i] = schema.Float(smax)
}
jobMetric.StatisticsSeries.Mean = mean
jobMetric.StatisticsSeries.Min = min
jobMetric.StatisticsSeries.Max = max
jobMetric.StatisticsSeries = &schema.StatsSeries{
Min: min, Mean: mean, Max: max,
}
jobMetric.Series = nil
}
}

View File

@ -7,6 +7,7 @@ import (
"encoding/json"
"errors"
"fmt"
"log"
"net/http"
"strconv"
"time"
@ -81,6 +82,7 @@ func (ccms *CCMetricStore) doRequest(job *schema.Job, suffix string, metrics []s
}
func (ccms *CCMetricStore) LoadData(job *schema.Job, metrics []string, scopes []schema.MetricScope, ctx context.Context) (schema.JobData, error) {
// log.Printf("job: %#v", job)
type ApiQuery struct {
Metric string `json:"metric"`
@ -106,7 +108,7 @@ func (ccms *CCMetricStore) LoadData(job *schema.Job, metrics []string, scopes []
reqBody := ApiQueryRequest{
Cluster: job.Cluster,
From: job.StartTime.Unix(),
To: job.StartTime.Add(time.Duration(job.Duration)).Unix(),
To: job.StartTime.Add(time.Duration(job.Duration) * time.Second).Unix(),
Queries: make([]ApiQuery, 0),
}
@ -118,12 +120,20 @@ func (ccms *CCMetricStore) LoadData(job *schema.Job, metrics []string, scopes []
scopeForMetric := map[string]schema.MetricScope{}
for _, metric := range metrics {
mc := config.GetMetricConfig(job.Cluster, metric)
if mc == nil {
// return nil, fmt.Errorf("metric '%s' is not specified for cluster '%s'", metric, job.Cluster)
log.Printf("metric '%s' is not specified for cluster '%s'", metric, job.Cluster)
continue
}
nativeScope, requestedScope := mc.Scope, scopes[0]
// case 1: A metric is requested at node scope with a native scope of node as well
// case 2: A metric is requested at node scope and node is exclusive
// case 3: A metric has native scope node
if (nativeScope == requestedScope && nativeScope == schema.MetricScopeNode) ||
(job.Exclusive == 1 && requestedScope == schema.MetricScopeNode) {
(job.Exclusive == 1 && requestedScope == schema.MetricScopeNode) ||
(nativeScope == schema.MetricScopeNode) {
nodes := map[string]bool{}
for _, resource := range job.Resources {
nodes[resource.Hostname] = true
@ -188,6 +198,8 @@ func (ccms *CCMetricStore) LoadData(job *schema.Job, metrics []string, scopes []
panic("todo")
}
// log.Printf("query: %#v", reqBody)
buf := &bytes.Buffer{}
if err := json.NewEncoder(buf).Encode(reqBody); err != nil {
return nil, err
@ -213,9 +225,16 @@ func (ccms *CCMetricStore) LoadData(job *schema.Job, metrics []string, scopes []
return nil, err
}
// log.Printf("response: %#v", resBody)
var jobData schema.JobData = make(schema.JobData)
for _, res := range resBody {
metric := res.Query.Metric
if _, ok := jobData[metric]; !ok {
jobData[metric] = make(map[schema.MetricScope]*schema.JobMetric)
}
if res.Error != nil {
return nil, fmt.Errorf("cc-metric-store error while fetching %s: %s", metric, *res.Error)
}
@ -239,6 +258,14 @@ func (ccms *CCMetricStore) LoadData(job *schema.Job, metrics []string, scopes []
*id, _ = strconv.Atoi(res.Query.TypeIds[0])
}
if res.Avg.IsNaN() || res.Min.IsNaN() || res.Max.IsNaN() {
// TODO: use schema.Float instead of float64?
// This is done because regular float64 can not be JSONed when NaN.
res.Avg = schema.Float(0)
res.Min = schema.Float(0)
res.Max = schema.Float(0)
}
jobMetric.Series = append(jobMetric.Series, schema.Series{
Hostname: res.Query.Hostname,
Id: id,

View File

@ -63,7 +63,13 @@ func LoadData(job *schema.Job, metrics []string, scopes []schema.MetricScope, ct
return nil, fmt.Errorf("no metric data repository configured for '%s'", job.Cluster)
}
return repo.LoadData(job, metrics, scopes, ctx)
data, err := repo.LoadData(job, metrics, scopes, ctx)
if err != nil {
return nil, err
}
calcStatisticsSeries(job, data)
return data, nil
}
data, err := loadFromArchive(job)

View File

@ -11,7 +11,6 @@ import (
// Common subset of Job and JobMeta. Use one of those, not
// this type directly.
type BaseJob struct {
ID int64 `json:"id" db:"id"`
JobID int64 `json:"jobId" db:"job_id"`
User string `json:"user" db:"user"`
Project string `json:"project" db:"project"`
@ -27,14 +26,15 @@ type BaseJob struct {
State JobState `json:"jobState" db:"job_state"`
Duration int32 `json:"duration" db:"duration"`
Tags []*Tag `json:"tags"`
RawResources []byte `json:"-" db:"resources"`
Resources []*Resource `json:"resources"`
MetaData interface{} `json:"metaData" db:"meta_data"`
}
// This type is used as the GraphQL interface and using sqlx as a table row.
type Job struct {
ID int64 `json:"id" db:"id"`
BaseJob
RawResources []byte `json:"-" db:"resources"`
StartTime time.Time `json:"startTime" db:"start_time"`
MemUsedMax float64 `json:"-" db:"mem_used_max"`
FlopsAnyAvg float64 `json:"-" db:"flops_any_avg"`
@ -52,7 +52,7 @@ type Job struct {
// the StartTime field with one of type int64.
type JobMeta struct {
BaseJob
StartTime int64 `json:"startTime"`
StartTime int64 `json:"startTime" db:"start_time"`
Statistics map[string]JobStatistics `json:"statistics,omitempty"`
}
@ -68,16 +68,6 @@ var JobColumns []string = []string{
"job.duration", "job.resources", "job.meta_data",
}
const JobInsertStmt string = `INSERT INTO job (
job_id, user, project, cluster, partition, array_job_id, num_nodes, num_hwthreads, num_acc,
exclusive, monitoring_status, smt, job_state, start_time, duration, resources, meta_data,
mem_used_max, flops_any_avg, mem_bw_avg, load_avg, net_bw_avg, net_data_vol_total, file_bw_avg, file_data_vol_total
) VALUES (
:job_id, :user, :project, :cluster, :partition, :array_job_id, :num_nodes, :num_hwthreads, :num_acc,
:exclusive, :monitoring_status, :smt, :job_state, :start_time, :duration, :resources, :meta_data,
:mem_used_max, :flops_any_avg, :mem_bw_avg, :load_avg, :net_bw_avg, :net_data_vol_total, :file_bw_avg, :file_data_vol_total
);`
type Scannable interface {
StructScan(dest interface{}) error
}
@ -85,7 +75,7 @@ type Scannable interface {
// Helper function for scanning jobs with the `jobTableCols` columns selected.
func ScanJob(row Scannable) (*Job, error) {
job := &Job{BaseJob: JobDefaults}
if err := row.StructScan(&job); err != nil {
if err := row.StructScan(job); err != nil {
return nil, err
}
@ -97,6 +87,7 @@ func ScanJob(row Scannable) (*Job, error) {
job.Duration = int32(time.Since(job.StartTime).Seconds())
}
job.RawResources = nil
return job, nil
}