Logs are difficult to understand after a 3-4 redeployments on the big env

[Sergey Galkin]

After 3-4 deployment on the cluster with 100 nodes we have logs that difficult to understand
On example we have node with 1 day uptime and 4 deployment
[root@fuel ~]# uptime
 16:46:05 up 1 day, 5:07

[root@fuel ~]# ls -l /var/log/remote/ | grep node-81
lrwxrwxrwx 1 root root 18 Nov 27 13:55 10.20.1.130 -> node-81.domain.tld
lrwxrwxrwx 1 root root 18 Nov 28 10:52 10.20.1.82 -> node-81.domain.tld
drwxr-xr-x 4 root root 4096 Nov 28 12:01 node-81.domain.tld
drwxr-xr-x 4 root root 4096 Nov 28 07:35 node-81.domain.tld.bak

[root@fuel ~]# du -sh /var/log/
11G /var/log/

We should have the way to separate logs (and snapshots) between deployments

[Aleksey Kasatkin (alekseyk-ru)]

There is a BP for this purpose: https://blueprints.launchpad.net/fuel/+spec/enable-per-environment-snapshots.

[Ihor Kalnytskyi (ikalnytskyi)]

It won't be easy to implement per environment snapshots. The problem is that we might have the same domain after few deployments, so the logs inside "node-81.domain.tld" folder will have logs from two envs: old one and new one.

[Alexander Bozhenko (alexbozhenko)]

Related bug
https://bugs.launchpad.net/fuel/+bug/1428825

[Vladimir Sharshov (vsharshov)]

Feature. Need bp. Moving to 8.0