I want to increase the response_header_timeout in thanos query-frontend deployment manifest as my queries are timing out, but can't find the correct syntax format expected.
Current Syntax I am following:
- -|
--query-frontend.downstream-tripper-config="config":
"response_header_timeout": "5m"
Other Syntax I have tried:
--query-frontend.downstream-tripper-config.response_header_timeout="5m"
But invalid.
Documentation Reference Link: Query-Frontend Flags
Example Manifest File: Example Manifest File
Example shown in docs: How to use config flags?
2.
Also there is a type mentioned in the example manifest file like:
type: "memcached"
What flag do I need to write in for query-frontend?
After editing the file, the deployment tried to use the edited file to create the pods but it sent the pod in CrashBackLoopOff state.
Thanks.
tl;dr: it will work in this format.
I had the same issue and based on the error log, it doesn't expect to receive the params nested under
"config":and fails to parse the yaml.Log example: (Line 1: field config not found)
level=error ts=2023-07-27T11:21:06.840839711Z caller=main.go:135 err="yaml: unmarshal errors:\n line 1: field config not found in type queryfrontend.DownstreamTripperConfig\nparsing downstream tripperI guess that
"config":is unnecessary here because DownstreamTripperConfig is already atype, as there are no different types so you just list all parameters there as they are.I found it here: https://github.com/thanos-io/thanos/blob/a817dbf586244bb7cbaa182bf830ad30b7e03944/pkg/queryfrontend/config.go#L181 It expects to read params directly from the provided yaml, e.g.,
yaml:"idle_conn_timeout"Comparing to CacheProviderConfig config, e.g. memcached, where it expects a different yaml format with
"config"and"type"https://github.com/thanos-io/thanos/blob/a817dbf586244bb7cbaa182bf830ad30b7e03944/pkg/queryfrontend/config.go#L84Hope it helps :)