| === mpjetta_ is now known as mpjetta | ||
| === mpjetta_ is now known as mpjetta | ||
| === mpjetta_ is now known as mpjetta | ||
| === Dmitrii-Sh-PTO is now known as Dmitrii-Sh | ||
| bobeo | hey everyone! | 19:36 |
|---|---|---|
| bobeo | we have a weird issue. after an abrupt power loss, the system we used to bootstrap and deploy our juju controller on no longer receives any forms of response back from the juju controller that was deployed. any ideas? | 19:37 |
| thumper | bobeo: did the controller lose power at all? | 20:19 |
| thumper | bobeo: also, which provider? | 20:20 |
| bobeo | @thumper | 20:24 |
| bobeo | thumper: Yes, all systems lost power, drives on some servers were swapped, a few broken, it was pretty bad. | 20:25 |
| thumper | bobeo: which provider are we talking about? | 20:26 |
| bobeo | thumper: its a private cloud. we use maas | 20:26 |
| thumper | In general, a reboot of a machine should break anything | 20:26 |
| thumper | even unplanned ones | 20:26 |
| thumper | however, swapped drives is something a bit different | 20:26 |
| thumper | if the controllers came up with different ip addresses, for example, that could break things | 20:27 |
| bobeo | thumper: from what we can see, the system that originally deployed the controller is still up. we still have access to the system, and it didnt have a drive swapped thankfully | 20:27 |
| bobeo | thumper: when it boots up, it says its starting up the juju database, which makes us think its still there, and the data and everything is still there | 20:27 |
| bobeo | thumper: but we dont know how to resync it back, or if thats even possible. | 20:27 |
| bobeo | thumper: sorry, the system that was originally bootstrapped up as a controller, the juju controller server, and the server used to deploy the juju controller, are both still up and accessibile | 20:28 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!