[MXS-1699] Improve log msg about the maxscale startup process Created: 2018-03-05 Updated: 2020-08-25 Resolved: 2018-03-08 |
|
| Status: | Closed |
| Project: | MariaDB MaxScale |
| Component/s: | Core |
| Affects Version/s: | 2.2.2 |
| Fix Version/s: | 2.1.15 |
| Type: | Bug | Priority: | Major |
| Reporter: | Nilnandan Joshi | Assignee: | markus makela |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Sprint: | MXS-SPRINT-53 |
| Description |
|
We need to improve log msg about the maxscale startup process so we can know how many listeners are started and how many are left and if remaining are failed then what was the reason? Recently, we saw one issue where maxscale was not starting and giving error like "error : [binlogrouter] Could not post ' ' message to main worker" The reason behind it was, while having multiple listeners, it was taking time to load all user data from the backend servers. The default value of TimeoutStartSec in maxscale.service for MaxScale is 120. If all listeners do not start before 120 seconds then it got timed out but we didn't get this information in maxscale log. More descriptive log message will help a lot. |
| Comments |
| Comment by Johan Wikman [ 2018-03-05 ] | |
|
So when TimeoutStartSec was increased MaxScale started and you do not see any error : [binlogrouter] Could not post ' ' message to main worker? Namely, there is no obvious connection between those two. | |
| Comment by markus makela [ 2018-03-05 ] | |
|
Added progress messages to the startup process. To figure out if there's something wrong with the messaging system, a separate issue for the following error should be created if it is reproducible.
|