Merge branch 'feat/565-add-metric-tooltip' into metric-store-tickets

Resolve conflicts in the generated GraphQL code by regenerating it against
the merged schema. cc-lib v2.13.0 adds Tooltip to schema.MetricConfig and
schema.GlobalMetricListItem, so gqlgen now binds the tooltip field directly
and the hand-written globalMetricListItem/metricConfig resolvers introduced
on the tooltip branch are no longer needed.

Also migrate to the cc-lib v2.13.0 metric container types, which changed
from bare maps to structs carrying array-valued metric groups:

  schema.JobData          map -> {Metrics, Groups}
  schema.ScopedJobStats   map -> {Metrics, Groups}
  job.Statistics          map -> schema.JobStatisticsSet{Metrics, Groups}

Callers index .Metrics, return the zero struct instead of nil, and
deepCopy/DecodeJobStats now also carry the Groups payload through.
archive.GetStatistics returns the full JobStatisticsSet so group
statistics survive the round trip.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-08-27 10:22:22 +02:00
co-authored by Claude Opus 5
35 changed files with 411 additions and 212 deletions
+38 -17
View File
@@ -117,7 +117,7 @@ func LoadData(job *schema.Job,
jd, err = ms.LoadData(job, metrics, scopes, ctx, resolution)
if err != nil {
if len(jd) != 0 {
if len(jd.Metrics) != 0 {
cclog.Warnf("partial error loading metrics from store for job %d (user: %s, project: %s, cluster: %s-%s): %s",
job.JobID, job.User, job.Project, job.Cluster, job.SubCluster, err.Error())
} else {
@@ -144,7 +144,7 @@ func LoadData(job *schema.Job,
if rfErr != nil {
return rfErr, 0, 0
}
for _, v := range jd {
for _, v := range jd.Metrics {
for _, v_ := range v {
timestep := int64(0)
for i := 0; i < len(v_.Series); i += 1 {
@@ -160,17 +160,20 @@ func LoadData(job *schema.Job,
// Filter job data to only include requested metrics and scopes, avoiding unnecessary data transfer.
if metrics != nil || scopes != nil {
if metrics == nil {
metrics = make([]string, 0, len(jd))
for k := range jd {
metrics = make([]string, 0, len(jd.Metrics))
for k := range jd.Metrics {
metrics = append(metrics, k)
}
}
res := schema.JobData{}
res := schema.JobData{
Metrics: make(map[string]schema.ScopedMetrics, len(metrics)),
Groups: jd.Groups,
}
for _, metric := range metrics {
if perscope, ok := jd[metric]; ok {
if perscope, ok := jd.Metrics[metric]; ok {
if len(perscope) > 1 {
subset := make(map[schema.MetricScope]*schema.JobMetric)
subset := make(schema.ScopedMetrics)
for _, scope := range scopes {
if jm, ok := perscope[scope]; ok {
subset[scope] = jm
@@ -182,7 +185,7 @@ func LoadData(job *schema.Job,
}
}
res[metric] = perscope
res.Metrics[metric] = perscope
}
}
jd = res
@@ -199,7 +202,7 @@ func LoadData(job *schema.Job,
// instead of overwhelming the UI with individual node lines. Note that newly calculated
// statistics use min/median/max, while archived statistics may use min/mean/max.
const maxSeriesSize int = 8
for _, scopes := range jd {
for _, scopes := range jd.Metrics {
for _, jm := range scopes {
if jm.StatisticsSeries != nil || len(jm.Series) < maxSeriesSize {
continue
@@ -229,7 +232,7 @@ func LoadData(job *schema.Job,
if err, ok := data.(error); ok {
cclog.Errorf("error in cached dataset for job %d: %s", job.JobID, err.Error())
return nil, err
return schema.JobData{}, err
}
return data.(schema.JobData), nil
@@ -296,14 +299,14 @@ func LoadScopedJobStats(
if err != nil {
cclog.Errorf("failed to access metricDataRepo for cluster %s-%s: %s",
job.Cluster, job.SubCluster, err.Error())
return nil, err
return schema.ScopedJobStats{}, err
}
scopedStats, err := ms.LoadScopedStats(job, metrics, scopes, ctx)
if err != nil {
cclog.Warnf("failed to load scoped statistics from metric store for job %d (user: %s, project: %s, cluster: %s-%s): %s",
job.JobID, job.User, job.Project, job.Cluster, job.SubCluster, err.Error())
return nil, err
return schema.ScopedJobStats{}, err
}
// Round Resulting Stat Values
@@ -451,7 +454,7 @@ func LoadNodeListData(
// Statistics are calculated as min/median/max.
const maxSeriesSize int = 8
for _, jd := range data {
for _, scopes := range jd {
for _, scopes := range jd.Metrics {
for _, jm := range scopes {
if jm.StatisticsSeries != nil || len(jm.Series) < maxSeriesSize {
continue
@@ -472,14 +475,32 @@ func LoadNodeListData(
// archived data (e.g., during resampling). This ensures the cached archive data remains
// immutable while allowing per-request transformations.
func deepCopy(source schema.JobData) schema.JobData {
result := make(schema.JobData, len(source))
result := schema.JobData{Metrics: copyScopedMetrics(source.Metrics)}
for _, group := range source.Groups {
copied := schema.MetricGroup{Key: group.Key}
for _, inst := range group.Instances {
copied.Instances = append(copied.Instances, schema.MetricGroupInstance{
Name: inst.Name,
Type: inst.Type,
Metrics: copyScopedMetrics(inst.Metrics),
})
}
result.Groups = append(result.Groups, copied)
}
return result
}
func copyScopedMetrics(source map[string]schema.ScopedMetrics) map[string]schema.ScopedMetrics {
result := make(map[string]schema.ScopedMetrics, len(source))
for metricName, scopeMap := range source {
result[metricName] = make(map[schema.MetricScope]*schema.JobMetric, len(scopeMap))
scopes := make(schema.ScopedMetrics, len(scopeMap))
for scope, jobMetric := range scopeMap {
result[metricName][scope] = copyJobMetric(jobMetric)
scopes[scope] = copyJobMetric(jobMetric)
}
result[metricName] = scopes
}
return result
+22 -22
View File
@@ -13,7 +13,7 @@ import (
func TestDeepCopy(t *testing.T) {
nodeId := "0"
original := schema.JobData{
original := schema.JobData{Metrics: map[string]schema.ScopedMetrics{
"cpu_load": {
schema.MetricScopeNode: &schema.JobMetric{
Timestep: 60,
@@ -42,42 +42,42 @@ func TestDeepCopy(t *testing.T) {
},
},
},
}
}}
copied := deepCopy(original)
original["cpu_load"][schema.MetricScopeNode].Series[0].Data[0] = 999.0
original["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Min[0] = 888.0
original["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles[25][0] = 777.0
original.Metrics["cpu_load"][schema.MetricScopeNode].Series[0].Data[0] = 999.0
original.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Min[0] = 888.0
original.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles[25][0] = 777.0
if copied["cpu_load"][schema.MetricScopeNode].Series[0].Data[0] != 1.0 {
if copied.Metrics["cpu_load"][schema.MetricScopeNode].Series[0].Data[0] != 1.0 {
t.Errorf("Series data was not deeply copied: got %v, want 1.0",
copied["cpu_load"][schema.MetricScopeNode].Series[0].Data[0])
copied.Metrics["cpu_load"][schema.MetricScopeNode].Series[0].Data[0])
}
if copied["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Min[0] != 1.0 {
if copied.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Min[0] != 1.0 {
t.Errorf("StatisticsSeries was not deeply copied: got %v, want 1.0",
copied["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Min[0])
copied.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Min[0])
}
if copied["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles[25][0] != 1.5 {
if copied.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles[25][0] != 1.5 {
t.Errorf("Percentiles was not deeply copied: got %v, want 1.5",
copied["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles[25][0])
copied.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles[25][0])
}
if copied["cpu_load"][schema.MetricScopeNode].Timestep != 60 {
if copied.Metrics["cpu_load"][schema.MetricScopeNode].Timestep != 60 {
t.Errorf("Timestep not copied correctly: got %v, want 60",
copied["cpu_load"][schema.MetricScopeNode].Timestep)
copied.Metrics["cpu_load"][schema.MetricScopeNode].Timestep)
}
if copied["cpu_load"][schema.MetricScopeNode].Series[0].Hostname != "node001" {
if copied.Metrics["cpu_load"][schema.MetricScopeNode].Series[0].Hostname != "node001" {
t.Errorf("Hostname not copied correctly: got %v, want node001",
copied["cpu_load"][schema.MetricScopeNode].Series[0].Hostname)
copied.Metrics["cpu_load"][schema.MetricScopeNode].Series[0].Hostname)
}
}
func TestDeepCopyNilStatisticsSeries(t *testing.T) {
original := schema.JobData{
original := schema.JobData{Metrics: map[string]schema.ScopedMetrics{
"mem_used": {
schema.MetricScopeNode: &schema.JobMetric{
Timestep: 60,
@@ -90,18 +90,18 @@ func TestDeepCopyNilStatisticsSeries(t *testing.T) {
StatisticsSeries: nil,
},
},
}
}}
copied := deepCopy(original)
if copied["mem_used"][schema.MetricScopeNode].StatisticsSeries != nil {
if copied.Metrics["mem_used"][schema.MetricScopeNode].StatisticsSeries != nil {
t.Errorf("StatisticsSeries should be nil, got %v",
copied["mem_used"][schema.MetricScopeNode].StatisticsSeries)
copied.Metrics["mem_used"][schema.MetricScopeNode].StatisticsSeries)
}
}
func TestDeepCopyEmptyPercentiles(t *testing.T) {
original := schema.JobData{
original := schema.JobData{Metrics: map[string]schema.ScopedMetrics{
"cpu_load": {
schema.MetricScopeNode: &schema.JobMetric{
Timestep: 60,
@@ -115,11 +115,11 @@ func TestDeepCopyEmptyPercentiles(t *testing.T) {
},
},
},
}
}}
copied := deepCopy(original)
if copied["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles != nil {
if copied.Metrics["cpu_load"][schema.MetricScopeNode].StatisticsSeries.Percentiles != nil {
t.Errorf("Percentiles should be nil when source is nil/empty")
}
}