Skip to content

feat: add prometheus metric for tracking user statuses #15281

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.

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

Merged
merged 5 commits into from
Oct 30, 2024
Merged
Show file tree
Hide file tree
Changes from 2 commits
Commits
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
8 changes: 7 additions & 1 deletion cli/server.go
Original file line number Diff line number Diff line change
Expand Up @@ -212,7 +212,13 @@ func enablePrometheus(
options.PrometheusRegistry.MustRegister(collectors.NewGoCollector())
options.PrometheusRegistry.MustRegister(collectors.NewProcessCollector(collectors.ProcessCollectorOpts{}))

closeUsersFunc, err := prometheusmetrics.ActiveUsers(ctx, options.PrometheusRegistry, options.Database, 0)
closeActiveUsersFunc, err := prometheusmetrics.ActiveUsers(ctx, options.Logger.Named("active_user_metrics"), options.PrometheusRegistry, options.Database, 0)
if err != nil {
return nil, xerrors.Errorf("register active users prometheus metric: %w", err)
}
afterCtx(ctx, closeActiveUsersFunc)

closeUsersFunc, err := prometheusmetrics.Users(ctx, options.Logger.Named("user_metrics"), options.PrometheusRegistry, options.Database, 0)
if err != nil {
return nil, xerrors.Errorf("register active users prometheus metric: %w", err)
}
Expand Down
54 changes: 53 additions & 1 deletion coderd/prometheusmetrics/prometheusmetrics.go
Original file line number Diff line number Diff line change
Expand Up @@ -27,7 +27,7 @@
const defaultRefreshRate = time.Minute

// ActiveUsers tracks the number of users that have authenticated within the past hour.
func ActiveUsers(ctx context.Context, registerer prometheus.Registerer, db database.Store, duration time.Duration) (func(), error) {
func ActiveUsers(ctx context.Context, logger slog.Logger, registerer prometheus.Registerer, db database.Store, duration time.Duration) (func(), error) {
if duration == 0 {
duration = defaultRefreshRate
}
Expand Down Expand Up @@ -58,6 +58,7 @@

apiKeys, err := db.GetAPIKeysLastUsedAfter(ctx, dbtime.Now().Add(-1*time.Hour))
if err != nil {
logger.Error(ctx, "get api keys", slog.Error(err))

Check failure on line 61 in coderd/prometheusmetrics/prometheusmetrics.go

View workflow job for this annotation

GitHub Actions / lint

ruleguard: Message "get api keys" is too short, it must be at least 16 characters long. (gocritic)
continue
}
distinctUsers := map[uuid.UUID]struct{}{}
Expand All @@ -73,6 +74,57 @@
}, nil
}

// Users tracks the total user count, broken out by status..
func Users(ctx context.Context, logger slog.Logger, registerer prometheus.Registerer, db database.Store, duration time.Duration) (func(), error) {
if duration == 0 {
// It's not super important this tracks real-time.
duration = defaultRefreshRate * 5
}

gauge := prometheus.NewGaugeVec(prometheus.GaugeOpts{
Namespace: "coderd",
Subsystem: "api",
Name: "total_user_count",
Help: "The total number of users, broken out by status.",
}, []string{"status"})
err := registerer.Register(gauge)
if err != nil {
return nil, err
}

ctx, cancelFunc := context.WithCancel(ctx)
done := make(chan struct{})
ticker := time.NewTicker(duration)
go func() {
defer close(done)
defer ticker.Stop()
for {
select {
case <-ctx.Done():
return
case <-ticker.C:
}

gauge.Reset()
//nolint:gocritic // This is a system service that needs full access
//to the users table.
users, err := db.GetUsers(dbauthz.AsSystemRestricted(ctx), database.GetUsersParams{})
if err != nil {
logger.Error(ctx, "get all users for prometheus metrics", slog.Error(err))
continue
}

for _, user := range users {
gauge.WithLabelValues(string(user.Status)).Inc()
}
}
}()
return func() {
cancelFunc()
<-done
}, nil
}

// Workspaces tracks the total number of workspaces with labels on status.
func Workspaces(ctx context.Context, logger slog.Logger, registerer prometheus.Registerer, db database.Store, duration time.Duration) (func(), error) {
if duration == 0 {
Expand Down
77 changes: 76 additions & 1 deletion coderd/prometheusmetrics/prometheusmetrics_test.go
Original file line number Diff line number Diff line change
Expand Up @@ -98,7 +98,7 @@ func TestActiveUsers(t *testing.T) {
t.Run(tc.Name, func(t *testing.T) {
t.Parallel()
registry := prometheus.NewRegistry()
closeFunc, err := prometheusmetrics.ActiveUsers(context.Background(), registry, tc.Database(t), time.Millisecond)
closeFunc, err := prometheusmetrics.ActiveUsers(context.Background(), slogtest.Make(t, nil), registry, tc.Database(t), time.Millisecond)
require.NoError(t, err)
t.Cleanup(closeFunc)

Expand All @@ -112,6 +112,81 @@ func TestActiveUsers(t *testing.T) {
}
}

func TestUsers(t *testing.T) {
t.Parallel()

for _, tc := range []struct {
Name string
Database func(t *testing.T) database.Store
Count map[string]int
}{{
Name: "None",
Database: func(t *testing.T) database.Store {
return dbmem.New()
},
Count: map[string]int{},
}, {
Name: "One",
Database: func(t *testing.T) database.Store {
db := dbmem.New()
dbgen.User(t, db, database.User{Status: database.UserStatusActive})
return db
},
Count: map[string]int{"active": 1},
}, {
Name: "MultipleStatuses",
Database: func(t *testing.T) database.Store {
db := dbmem.New()

dbgen.User(t, db, database.User{Status: database.UserStatusActive})
dbgen.User(t, db, database.User{Status: database.UserStatusDormant})

return db
},
Count: map[string]int{"active": 1, "dormant": 1},
}, {
Name: "MultipleActive",
Database: func(t *testing.T) database.Store {
db := dbmem.New()
dbgen.User(t, db, database.User{Status: database.UserStatusActive})
dbgen.User(t, db, database.User{Status: database.UserStatusActive})
dbgen.User(t, db, database.User{Status: database.UserStatusActive})
return db
},
Count: map[string]int{"active": 3},
}} {
tc := tc
t.Run(tc.Name, func(t *testing.T) {
t.Parallel()
registry := prometheus.NewRegistry()
closeFunc, err := prometheusmetrics.Users(context.Background(), slogtest.Make(t, nil), registry, tc.Database(t), time.Millisecond)
require.NoError(t, err)
t.Cleanup(closeFunc)

require.Eventually(t, func() bool {
metrics, err := registry.Gather()
if err != nil {
return false
}

// If we get no metrics and we know none should exist, bail
// early. If we get no metrics but we expect some, retry.
if len(metrics) == 0 {
return len(tc.Count) == 0
}

for _, metric := range metrics[0].Metric {
if tc.Count[*metric.Label[0].Value] != int(metric.Gauge.GetValue()) {
return false
}
}

return true
}, testutil.WaitShort, testutil.IntervalSlow)
})
}
}

func TestWorkspaceLatestBuildTotals(t *testing.T) {
t.Parallel()

Expand Down
Loading