netbird/gvisor
0
0
Fork 0
mirror of https://github.com/netbirdio/gvisor.git synced 2026-05-22 17:12:49 -07:00
Code Issues Packages Projects Releases Wiki Activity

runsc metric-server: Add per-sandbox start timestamp metric.

Browse Source 
This synthetic metric contains the Unix timestamp that each sandbox was
started at.

This is useful for counter metrics, such that rates of change over time can
be properly on a per-sandbox basis.

PiperOrigin-RevId: 505228878
This commit is contained in:
Etienne Perot
2023-01-27 15:56:37 -08:00
committed by gVisor bot
parent 87b3c88c32
commit 4e75dc4650
2 changed files with 28 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files
pkg/prometheus/prometheus.go
+7 -2
View File
@@ -222,7 +222,7 @@ type Data struct {
// NewIntData returns a new Data struct with the given metric and value.
func NewIntData(metric *Metric, val int64) *Data {
return &Data{Metric: metric, Number: &Number{Int: val}}
return LabeledIntData(metric, nil, val)
}
// LabeledIntData returns a new Data struct with the given metric, labels, and value.
@@ -232,7 +232,12 @@ func LabeledIntData(metric *Metric, labels map[string]string, val int64) *Data {
// NewFloatData returns a new Data struct with the given metric and value.
func NewFloatData(metric *Metric, val float64) *Data {
return &Data{Metric: metric, Number: &Number{Float: val}}
return LabeledFloatData(metric, nil, val)
}
// LabeledFloatData returns a new Data struct with the given metric, labels, and value.
func LabeledFloatData(metric *Metric, labels map[string]string, val float64) *Data {
return &Data{Metric: metric, Labels: labels, Number: &Number{Float: val}}
}
// ExportOptions contains options that control how metric data is exported in Prometheus format.
runsc/cmd/metric_server.go
+21 -15
View File
@@ -475,6 +475,11 @@ var (
Type: prometheus.TypeGauge,
Help: "Key-value pairs about per-sandbox metadata.",
}
sandboxCreationMetric = prometheus.Metric{
Name: "sandbox_creation_time_seconds",
Type: prometheus.TypeGauge,
Help: "When the sandbox was created, as a unix timestamp in milliseconds.",
}
numRunningSandboxesMetric = prometheus.Metric{
Name: "num_sandboxes_running",
Type: prometheus.TypeGauge,
@@ -595,12 +600,13 @@ func (m *MetricServer) serveMetrics(w http.ResponseWriter, req *http.Request) ht
go func(metricsMu *sync.Mutex, meta *metaMetrics, selfMetrics *prometheus.Snapshot) {
defer wg.Done()
for s := range loadedSandboxCh {
served, sand, verifier, err := s.served, s.sandbox, s.verifier, s.err
served, sand, verifier, loadErr := s.served, s.sandbox, s.verifier, s.err
isRunning := false
var snapshot *prometheus.Snapshot
if err == nil {
sandboxErr := loadErr
if loadErr == nil {
queryCtx, queryCtxCancel := context.WithTimeout(ctx, perSandboxTime)
snapshot, err = queryMetrics(queryCtx, sand, verifier)
snapshot, sandboxErr = queryMetrics(queryCtx, sand, verifier)
queryCtxCancel()
isRunning = sand.IsRunning()
}
@@ -608,25 +614,25 @@ func (m *MetricServer) serveMetrics(w http.ResponseWriter, req *http.Request) ht
metricsMu.Lock()
defer metricsMu.Unlock()
selfMetrics.Add(prometheus.LabeledIntData(&sandboxPresenceMetric, served.extraLabels, 1))
selfMetrics.Add(prometheus.LabeledIntData(&sandboxMetadataMetric, served.labelsWithMetadata, 1))
sandboxRunning := int64(0)
if isRunning {
sandboxRunning = 1
meta.numRunningSandboxes++
}
selfMetrics.Add(prometheus.LabeledIntData(&sandboxRunningMetric, served.extraLabels, sandboxRunning))
if err != nil && !isRunning {
// The sandbox either hasn't started running yet, or it ran and has gone away between the
// start of the function and now. It is normal that metrics are not exported for this
// sandbox in this case, so do not report this as an error.
if loadErr == nil {
selfMetrics.Add(prometheus.LabeledIntData(&sandboxMetadataMetric, served.labelsWithMetadata, 1))
selfMetrics.Add(prometheus.LabeledFloatData(&sandboxCreationMetric, served.extraLabels, float64(served.createdAt.Unix())+(float64(served.createdAt.Nanosecond())/1e9)))
}
if sandboxErr != nil {
// If the sandbox isn't running, it is normal that metrics are not exported for it, so
// do not report this case as an error.
if isRunning {
meta.numCannotExportSandboxes++
log.Warningf("Could not export metrics from sandbox %s: %v", served.rootContainerID.SandboxID, sandboxErr)
}
return
}
if err != nil {
meta.numRunningSandboxes++
meta.numCannotExportSandboxes++
log.Warningf("Could not export metrics from sandbox %s: %v", served.rootContainerID.SandboxID, err)
return
}
meta.numRunningSandboxes++
snapshotCh <- snapshotAndOptions{
snapshot: snapshot,
options: prometheus.SnapshotExportOptions{