These changes configure the remote read concurrency limit to a default value of 10.
The value can be configured / or the default can be changed if needed.
This limit defines how many remote read requests can be handled at the same time.
If not set this is 0 by default and remote read requests end in an timeout. (Since they are stuck in the gate for remote reads and waiting endlessly).
After setting this setting a prometheus remote read client can successfully use the remote read API of graphite-clickhouse.
For faster read support Streamed Chunks should be used with the chunk querier.
I have a prototype implementation for this, but cannot test it atm since prometheus does support streamed chunks as a client yet....
Will open a PR for this as soon as I have smth to test this on.
These changes configure the remote read concurrency limit to a default value of 10. The value can be configured / or the default can be changed if needed.
This limit defines how many remote read requests can be handled at the same time. If not set this is 0 by default and remote read requests end in an timeout. (Since they are stuck in the gate for remote reads and waiting endlessly).
After setting this setting a prometheus remote read client can successfully use the remote read API of graphite-clickhouse.
https://prometheus.io/docs/prometheus/latest/querying/remote_read_api/
Only samples are supported.
For faster read support Streamed Chunks should be used with the chunk querier. I have a prototype implementation for this, but cannot test it atm since prometheus does support streamed chunks as a client yet....
Will open a PR for this as soon as I have smth to test this on.