feat: add prometheus metric for tracking user statuses #15281
Conversation
| }, testutil.WaitShort, testutil.IntervalSlow) | ||
| } | ||
|
|
||
| require.Eventually(t, checkFn, testutil.WaitShort, testutil.IntervalFast) |
There was a problem hiding this comment.
Hhmm, do we need require.Eventually with quartz now implemented (thank you for doing this!)?
There was a problem hiding this comment.
It was my assumption that AdvanceNext().MustWait() only waits for the ticker to fire, then the actual gathering code runs async. I guess I could technically make it block until the end using AfterFunc, but that felt kinda janky. My assumption could definitely be wrong tho, in my testing the metrics seemed to always be ready on the first try but I didn't want to risk it.
| ctx, cancelFunc := context.WithCancel(ctx) | ||
| done := make(chan struct{}) | ||
| ticker := clk.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() | ||
| } | ||
| } |
There was a problem hiding this comment.
Instead of doing this you can just use clock.TickerFunc, it removes any timing problems in the tests as well
There was a problem hiding this comment.
Ah, that makes more sense
Adds the ability to track total users in the platform by status. This can help for seeing how license seat usage is being used over time.
Closes #15278