Resolving the "DEADLINE_EXCEEDED" error
Written by
Updated at November 27, 2023
Issue description
When attempting to execute this curl request:
curl "https://monitoring.api.cloud.yandex.net/monitoring/v2/prometheusMetrics/?folderId=folder_id&service=managed-kubernetes" -H "Authorization: Bearer ${IAM_TOKEN}" -H "Content-Type: application/json" -i
you see this error message:
{"code":504,"message":"Metabase: [cloud-prod-sas] Status{code=DEADLINE_EXCEEDED, description=deadline exceeded after 0.807915461s, ... cause=null}"}
Solution
This error arises because of numerous metrics, due to which the request does not have time to complete. We recommend exporting metrics using selectors to reduce the volume of data:
scrape_configs:
- job_name: 'yc-monitoring-export'
metrics_path: '/monitoring/v2/prometheusMetrics'
params:
folderId:
- 'b1gXXXXXXXXXXXXXXXXX'
service:
- 'managed-kubernetes'
selectors:
- 'namespace=...'
bearer_token: ...
Note
The selectors parameter for pod metrics can be filtered by namespace or pod name mask, for example, pod=part-of-name*. We write more details in the query language documentation.