boblack_zocker
boblack_zocker4mo ago

Performance breaks by unknown reasons

Hello, our production self-hosted zitadel is doing misterious things in the kubernetes. Out of nowhere he start big read/write operations and by that it stops working for about 5minutes till the process is done and then it functions normally again. is it doing some kind of maintance in the background or why is it doing that?
No description
No description
8 Replies
Raccine
Raccine4mo ago
Hey @boblack_zocker! Thanks for reaching out - Let me loop in an engineer to look into this for you! ☺️
boblack_zocker
boblack_zockerOP4mo ago
Thanks, is it possible to get an answer today? our production is having some outages because of this :). thanks
Rajat Singh
Rajat Singh4mo ago
@boblack_zocker apologies for what you are facing we are still looking into this , we will get back to you.
adlerhurst
adlerhurst4mo ago
Do you see what statements are executed?
boblack_zocker
boblack_zockerOP4mo ago
@adlerhurst what do you mean? i have some errorlogs but not more its running in its own pod in kubernetes it started with this logs: 2025-02-24 11:10:56.494 time="2025-02-24T10:10:56Z" level=info msg="process events failed" caller="/home/runner/work/zitadel/zitadel/internal/eventstore/handler/v2/handler.go:413" error="context deadline exceeded" projection=projections.instance_domains2025-02-24 11:10:56.310
time="2025-02-24T10:10:56Z" level=info msg="process events failed" caller="/home/runner/work/zitadel/zitadel/internal/eventstore/handler/v2/handler.go:413" error="context deadline exceeded" projection=projections.device_auth_requests2 2025-02-24 11:10:56.096
time="2025-02-24T10:10:55Z" level=info msg="query failed" caller="/home/runner/work/zitadel/zitadel/internal/eventstore/repository/sql/query.go:129" error="timeout: context already done: context deadline exceeded" 2025-02-24 11:10:56.074
time="2025-02-24T10:10:55Z" level=info msg="query failed" caller="/home/runner/work/zitadel/zitadel/internal/eventstore/repository/sql/query.go:129" error="context deadline exceeded" is this a problem with zitadel or the database of zitadel?
adlerhurst
adlerhurst4mo ago
I mean from the database some query insights. Can you share your zitadel config, im mostly interested in the projections part
boblack_zocker
boblack_zockerOP4mo ago
we got the default.yaml zitadel config what exacly do you need to see on the query insights?
adlerhurst
adlerhurst4mo ago
it looks to me like the statements take too long. What you can try is to set the following variables: to disable the timeouts
Projections:
TransactionDuration: 0
Auth:
Spooler:
TransactionDuration: 0 #ZITADEL_AUTH_SPOOLER_TRANSACTIONDURATION

Admin:
Spooler:
TransactionDuration: 10s
Projections:
TransactionDuration: 0
Auth:
Spooler:
TransactionDuration: 0 #ZITADEL_AUTH_SPOOLER_TRANSACTIONDURATION

Admin:
Spooler:
TransactionDuration: 10s

Did you find this page helpful?