You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Expected behavior:
If you restart the Postgres server which is providing Teleport's backend database, Teleport should reconnect and recover gracefully.
Current behavior:
Restarting the Postgres server doesn't appear to cause any errors, but attempting to perform an SSO login via Github resulted in an error:
It's possible that this would have resolved itself if I'd waited a little longer (this was probably ~20-30 seconds after the Postgres server restarted) but it seemed odd that only certain operations were erroring. Everything was fine after I restarted Teleport.
Bug details:
Teleport version: 9.2.3
The text was updated successfully, but these errors were encountered:
This appears to be an issue with the postgres driver we're using: jackc/pgx#672
I know lib/pq will receive the connection close from postgres' administrative broadcast on shutdown, store the error in the connection object in the connection pool, and on next attempt to execute a command will simply attempt to reconnect. Reading through that issue, it appears that pgx still lacks that capability in a stable release.
Expected behavior:
If you restart the Postgres server which is providing Teleport's backend database, Teleport should reconnect and recover gracefully.
Current behavior:
Restarting the Postgres server doesn't appear to cause any errors, but attempting to perform an SSO login via Github resulted in an error:
It's possible that this would have resolved itself if I'd waited a little longer (this was probably ~20-30 seconds after the Postgres server restarted) but it seemed odd that only certain operations were erroring. Everything was fine after I restarted Teleport.
Bug details:
The text was updated successfully, but these errors were encountered: