Polish logrotate configuration
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Committed
|
High
|
Bartłomiej Piotrowski | ||
6.0.x |
Won't Fix
|
High
|
Alexey Stupnikov |
Bug Description
This bug is continuation of https:/
1. Logrotate needs to be configured to skip our configuration file when running daily schedule. This can be achieved by 'tabooext + .ignoredaily' + conf file rename.
2. We should use 'su root syslog' on Ubuntu to prevent https:/
3. On Ubuntu additional directory, /var/log/upstart/, exists. Files put there aren't covered by our logrotate configuration. We can also tell upstart to stop logging there: https:/
4. Additionally logrotate can take even 45 minutes for rotation and compression. Only few files on controller nodes grow significantly faster, but it seems pointless to create separate config just for them.
5. Our custom script is put in the wrong directory, as seen here:
<75>Apr 29 02:58:01 node-1 cron[7074]: Error: bad minute; while reading /etc/cron.
<78>Apr 29 02:58:01 node-1 cron[7074]: (*system*
Changed in fuel: | |
status: | Triaged → In Progress |
Changed in fuel: | |
milestone: | 7.0 → 6.1 |
importance: | Medium → High |
tags: | added: customer-found |
tags: | added: support |
Bartlomiej, thank you for creating this bug.
I think Importance should be set to critical, and milestone to 6.1:
We can't afford to ship partially-working logrotate.