Replies: 9 comments
-
@kevinchalet could you please comment? |
Beta Was this translation helpful? Give feedback.
-
It could certainly run less frequently. As to why it's so expensive, it would be nice to profile it to see what's causing the spike (or we could just ignore that and wait for the secrets module to ship to replace all that stuff 🤣) |
Beta Was this translation helpful? Give feedback.
-
For reference: #7891 |
Beta Was this translation helpful? Give feedback.
-
Thank you for getting back to me, I'll keep an eye on #7891 . We've reduced the spikes by only allowing a single instance to handle the task for each tenant every 30 minutes, so we're not in a rush. But it would be nice to reduce the spikes even more. |
Beta Was this translation helpful? Give feedback.
-
If this task doesn't need to be executed by every tenant, like if it doesn't require tenant's state, it could be a host task. Or maybe this should be a feature in the Default tenant only. |
Beta Was this translation helpful? Give feedback.
-
Sadly, the logic it runs is tenant-aware, so unless we decide to make the OpenID server feature only usable in the default tenant, it won't work 😃 |
Beta Was this translation helpful? Give feedback.
-
Yes, it definitely does need to run for every tenant - just not multiple times per tenant, as it does when there are multiple instances. |
Beta Was this translation helpful? Give feedback.
-
Yes we plan to implement exclusive execution, for now we can only prevent concurrent execution through the Admin UI (need the RedisLock feature), see #10140 (comment). But in the meantime you can implement a full ownership, see #11463 (comment). |
Beta Was this translation helpful? Give feedback.
-
Yes, we had ended up going with a solution similar to the one in your second link. Thanks! |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Related to #13262 and #13109
Is there a reason that OpenIdBackgroundTask is set to run every 30 minutes? We get a large spike of certificate calls every 30 minutes because of this task, and have been trying to manage that. We thought it might help to run the task less often, but I don't know if that will just mean even bigger spikes when it does run, lagging, etc.
Do you think we could run it less frequently without too much risk? Thanks.
Beta Was this translation helpful? Give feedback.
All reactions