move of keepalived VIP address causes several minutes of cloud instability
Bug #1865843 reported by
Radosław Piliszek
This bug affects 3 people
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
kolla-ansible |
New
|
High
|
Unassigned |
Bug Description
If keepalived VIP address moves to another host, e.g. due to kolla-ansible run restarting keepalived container or due to any other reason, there is a lot of havoc in cloud, mostly due to MariaDB dropped connections. Random 500/504 errors happen for several minutes. Horizon is hit most harsh as it is pretty unusable during that period.
Logged errors are diverse but they have roots in these: "Lost connection to MySQL server during query" "Database connection was found disconnected"
Changed in kolla-ansible: | |
importance: | Undecided → High |
To post a comment you must log in.
Just as reference, I've asked in the IRC channel about this bug and yoctozepto mentioned two patches have been proposed about it: /review. opendev. org/#/c/ 723342/ /review. opendev. org/#/c/ 749632/
- https:/
- https:/