Skip to content

Surface failed ECS API requests to the collector #111

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Surface failed ECS API requests to the collector #111

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 19 additions & 2 deletions ecscollector/collector.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@ package ecscollector

import (
"context"
"fmt"
"log/slog"

"github.com/docker/docker/api/types/container"
Expand Down Expand Up @@ -193,7 +194,15 @@ func (c *collector) Collect(ch chan<- prometheus.Metric) {
ctx := context.Background()
metadata, err := c.client.RetrieveTaskMetadata(ctx)
if err != nil {
c.logger.Debug("Failed to retrieve metadata", "error", err)
c.logger.Debug("Failed to retrieve task metadata", "error", err)
// Signal that this Collect has failed. This ultimately results in an
// HTTP 500 being served for the /metrics response.
//
// While it would be most technically correct to do `NewInvalidMetric`
// for all of the Descs here, it just results in N identical error
// messages being printed in the /metrics response, so it seems a bit
// absurd to do that.
ch <- prometheus.NewInvalidMetric(taskMetadataDesc, fmt.Errorf("failed to retrieve task metadata: %w", err))
return
}
c.logger.Debug("Got ECS task metadata response", "metadata", metadata)
Expand Down Expand Up @@ -261,7 +270,15 @@ func (c *collector) Collect(ch chan<- prometheus.Metric) {

stats, err := c.client.RetrieveTaskStats(ctx)
if err != nil {
c.logger.Debug("Failed to retrieve container stats", "error", err)
c.logger.Debug("Failed to retrieve task stats", "error", err)
// Signal that this Collect has failed. This ultimately results in an
// HTTP 500 being served for the /metrics response.
//
// While it would be most technically correct to do `NewInvalidMetric`
// for all of the Descs here, it just results in N identical error
// messages being printed in the /metrics response, so it seems a bit
// absurd to do that.
ch <- prometheus.NewInvalidMetric(memUsageDesc, fmt.Errorf("failed to retrieve task stats: %w", err))
return
}
c.logger.Debug("Got ECS task stats response", "stats", stats)
Expand Down
21 changes: 21 additions & 0 deletions ecscollector/collector_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -138,3 +138,24 @@ func TestEc2Metrics(t *testing.T) {
collector := NewCollector(metadataClient, slog.Default())
assertSnapshot(t, collector, "testdata/snapshots/ec2_metrics.txt")
}

func TestApiErrors(t *testing.T) {
mux := http.NewServeMux()
mux.HandleFunc("GET /task", func(w http.ResponseWriter, r *http.Request) {
w.WriteHeader(500)
w.Write([]byte("Internal Server Error"))
})
mux.HandleFunc("GET /task/stats", func(w http.ResponseWriter, r *http.Request) {
w.WriteHeader(500)
w.Write([]byte("Internal Server Error"))
})
metadataServer := httptest.NewServer(mux)
defer metadataServer.Close()
metadataClient := ecsmetadata.NewClient(metadataServer.URL)
collector := NewCollector(metadataClient, slog.Default())

_, err := renderMetrics(collector)
if err == nil || err.Error() != "non-200 metrics response: 500" {
t.Fatalf("expected 500 error but got err: %v", err)
}
}
Loading