I ran into two issues with headless auth. The first one manifested itself by tsh hanging when waiting for approval after approving a headless auth request through the Web UI. The second one was Connect not showing a modal when a new request is created.
I figured that the issue was related to some problems with watchers in the auth service. @gzdunek reported running into the same problem. Then a minute later I get a message from him that restarting teleport helped. I restarted my cluster and everything started working again.
Searching for "headless" in the logs doesn't bring up anything interesting, though I didn't have debug logs enabled. Searching for "watcher" shows that there's a ton of the following logs. idk if they're related to headless auth.
2024-06-10T12:11:15+02:00 WARN [REVERSE:M] Re-init the cache on error: watcher is closed: error reading from server: EOF cache/cache.go:1131
Any further tips on how to debug this when it happens again?
I ran into two issues with headless auth. The first one manifested itself by tsh hanging when waiting for approval after approving a headless auth request through the Web UI. The second one was Connect not showing a modal when a new request is created.
I figured that the issue was related to some problems with watchers in the auth service. @gzdunek reported running into the same problem. Then a minute later I get a message from him that restarting
teleport
helped. I restarted my cluster and everything started working again.Searching for "headless" in the logs doesn't bring up anything interesting, though I didn't have debug logs enabled. Searching for "watcher" shows that there's a ton of the following logs. idk if they're related to headless auth.
Any further tips on how to debug this when it happens again?