Change Details

We should determine how #nasqueron-ops should work The initial proposal was: **A server log** Any action done on a server should be logged by an entry on the channel. The Phabricator bot (hmmm hmmm) should then log the action, so it's available in Chatlog for reference. This evening, we had a lot of operations issue, T710 T711 T712 T714 T715 T716 T717 T718 T719 and T720 but virtually nothing were logged on the channel. **A notifications channel** It were intended any incident, unresponsive server were immediately notified on the channel. Yet, @Kaliiixx noted we shouldn't clutter the channel with Phabricator notifications flood, so we can discuss on the channel. Bilan: we have notifications in the void and we don't know when there are new Docker images rebuilt because upstream images has changed. Or when @amj requested a SSH access, if I weren't there, the task wouldn't have been prioritized either, as no notification were done. **An operations work channel** It were also intended to discuss, organize operations work. I don't know how notifications could affect this part. ____ So we should do to find a balance between being correctly notified of what happens and be able to work on the channel in the same time. And in an ideal world, every notification should trigger an action, like handle the task or check the Docker image.

We should determine how #nasqueron-ops should work The initial proposal was: **A server log** Any action done on a server should be logged by an entry on the channel. The Phabricator bot (hmmm hmmm) should then log the action, so it's available in Chatlog for reference. Yesterday and this evening, we had a lot of operations issue, T710 T711 T712 T714 T715 T716 T717 T718 T719 and T720 but virtually nothing were logged on the channel. **A notifications channel** It were intended any incident, unresponsive server were immediately notified on the channel. Yet, @Kaliiixx noted we shouldn't clutter the channel with Phabricator notifications flood, so we can discuss on the channel. Bilan: we have notifications in the void and we don't know when there are new Docker images rebuilt because upstream images has changed. Or when @amj requested a SSH access, if I weren't there, the task wouldn't have been prioritized either, as no notification were done. **An operations work channel** It were also intended to discuss, organize operations work. I don't know how notifications could affect this part. ____ So we should do to find a balance between being correctly notified of what happens and be able to work on the channel in the same time. And in an ideal world, every notification should trigger an action, like handle the task or check the Docker image.

We should determine how #nasqueron-ops should work The initial proposal was: **A server log** Any action done on a server should be logged by an entry on the channel. The Phabricator bot (hmmm hmmm) should then log the action, so it's available in Chatlog for reference. TYesterday and this evening, we had a lot of operations issue, T710 T711 T712 T714 T715 T716 T717 T718 T719 and T720 but virtually nothing were logged on the channel. **A notifications channel** It were intended any incident, unresponsive server were immediately notified on the channel. Yet, @Kaliiixx noted we shouldn't clutter the channel with Phabricator notifications flood, so we can discuss on the channel. Bilan: we have notifications in the void and we don't know when there are new Docker images rebuilt because upstream images has changed. Or when @amj requested a SSH access, if I weren't there, the task wouldn't have been prioritized either, as no notification were done. **An operations work channel** It were also intended to discuss, organize operations work. I don't know how notifications could affect this part. ____ So we should do to find a balance between being correctly notified of what happens and be able to work on the channel in the same time. And in an ideal world, every notification should trigger an action, like handle the task or check the Docker image.