-
Notifications
You must be signed in to change notification settings - Fork 308
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
"500 Internal Server Error" because of "Duplicate entry" after Upgrade to PI-3.9.1 #3810
Comments
It seems to be related to this commit: And I understand that this commit should actually fix such problems. Strange. |
Hi @beckerr-rzht You have always only one PI node writing to the db? So this error happens during a failover? |
Good questions. ClusterThere are three VMs with IDP (Shibboleth), PI and Mariadb Galera nodes (1:10.3.38-0ubuntu0.20.04.1). Before the update to PI 3.9.1:
Updating PIOur seamess and quite well tested update procedure currently takes about 5 minutes and looks like this:
Result
ErrorsThe first errors occurred in this state immediately after the update to PI 3.9.1.
|
hey @plettich did we fix this already? maybe in 3.9.3 or .2? |
We operate PrivacyIdea in production with a MariaDB Galera cluster consisting of 3 nodes.
Each database node is assigned a PI node, which communicates with the local MariaDB node (on localhost).
One of the nodes is selected as the master via keepalived.
This worked without any problems up to PI 3.8.1. Especially when the master was changed via keepalive.
However, after upgrading from PI-3.8.1 to PI-3.9.1, I now see many "duplicate key" errors.
Example:
This should actually prevented by the Galera cluster, as its configuration regarding autoincrement is correct:
https://mariadb.com/kb/en/auto_increment/#replication
I have compared dumps of the PI database and see that many SEQUENCES have been added and suspect the cause of the problem there.
Example:
Since we first encountered the problem in the productive environment, I have this urgent question:
Is there a quick fix or workaround that gets around this problem?
The text was updated successfully, but these errors were encountered: