| *** suzhengwei__ has joined #openstack-masakari | 00:47 | |
| *** suzhengwei__ has quit IRC | 01:02 | |
| *** suzhengwei__ has joined #openstack-masakari | 01:02 | |
| *** suzhengwei__ has quit IRC | 01:32 | |
| *** suzhengwei__ has joined #openstack-masakari | 02:03 | |
| *** suzhengwei__ has quit IRC | 04:17 | |
| *** suzhengwei__ has joined #openstack-masakari | 05:31 | |
| *** osmanlic- has joined #openstack-masakari | 06:45 | |
| *** osmanlicilegi has quit IRC | 06:45 | |
| *** osmanlic- has quit IRC | 07:01 | |
| *** osmanlicilegi has joined #openstack-masakari | 07:05 | |
| *** suzhengwei__ has quit IRC | 09:11 | |
| *** suzhengwei__ has joined #openstack-masakari | 09:11 | |
| *** suzhengwei__ has quit IRC | 11:37 | |
| *** suzhengwei__ has joined #openstack-masakari | 11:37 | |
| *** suzhengwei__ has quit IRC | 11:43 | |
| *** suzhengwei__ has joined #openstack-masakari | 11:43 | |
| *** gwaewion has joined #openstack-masakari | 11:47 | |
| *** suzhengwei__ has quit IRC | 12:38 | |
| gwaewion | hi. does anyone use consul (pf9-ha) instead of corosync/pacemaker? if not, how many hosts may exists in one masakari segment? | 12:57 |
|---|---|---|
| yoctozepto | gwaewion: hi! there is no consul support in masakari at the moment; masakari with pacemaker remotes has been known to work with about a hundred compute nodes | 13:10 |
| yoctozepto | theoretically it can be scaled to several hundreds with the right setup and size of pacemaker cluster | 13:11 |
| yoctozepto | there is a goal to implement consul support in masakari to avoid the complexi of the pacemaker stack | 13:12 |
| yoctozepto | but we are not aware of any scalability/robustness issues arising from the use of pacemaker | 13:12 |
| yoctozepto | "to avoid the complexity* of" | 13:13 |
| gwaewion | yoctozepto: thank you. so, if i need more than 16 hosts in cluster (segment) i must use pacemaker-remote and do not pay attention to it size? | 13:52 |
| yoctozepto | gwaewion: you should probably already start using remotes before you extend the base cluster beyond 3 nodes | 13:53 |
| yoctozepto | it would be more efficient this way | 13:53 |
| gwaewion | have you heard about this thing - https://github.com/platform9/pf9-ha? i can not find any documentation nor detailed information about this stuff =( | 13:53 |
| yoctozepto | no, I have not | 13:53 |
| gwaewion | at the moment i have 12 nodes in cluster with "classic" pacemaker and every try of expansion can not be finished without killing (some time twice) pacemaker processes on every node in cluster | 13:56 |
| yoctozepto | oh my, I suggest you recreate the cluster with 3 primary nodes | 13:58 |
| yoctozepto | and add the other 9 ones as remotes | 13:58 |
| gwaewion | thank for advice, i will think about this | 14:01 |
| gwaewion | could you tell where i can read about state of consul support in masakari? | 14:01 |
| yoctozepto | gwaewion: it's non-existent in the merged code at the moment | 14:04 |
| yoctozepto | but you can read the spec | 14:04 |
| yoctozepto | https://review.opendev.org/c/openstack/masakari-specs/+/734017 | 14:04 |
| yoctozepto | and the proposed code | 14:04 |
| yoctozepto | https://review.opendev.org/c/openstack/masakari-monitors/+/785572 | 14:04 |
| yoctozepto | there is a chance we get something in during the upcoming 3 months | 14:05 |
| gwaewion | and again thank you for your help. will hope this will happen and i will not have to update to next openstack release =) | 14:15 |
| yoctozepto | well, it will be in some next openstack release anyway; features don't get backported to the stable branches | 14:26 |
| gwaewion | /me becomes sad because of this | 14:33 |
| *** suzhengwei__ has joined #openstack-masakari | 15:31 | |
| *** gwaewion has quit IRC | 15:59 | |
| *** suzhengwei__ has quit IRC | 16:01 | |
| *** gwaewion has joined #openstack-masakari | 16:15 | |
| *** gwaewion has quit IRC | 20:20 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!