Don't query all ingesters #1447
Labels
component/loki
keepalive
An issue or PR that will be kept alive and never marked as stale.
type/enhancement
Something existing could be improved
Comments
rfratto
added
component/loki
type/enhancement
cyriltovena
added
the
keepalive
rfratto
added a commit
to rfratto/loki
that referenced
this issue
Jan 6, 2020
This commit uses replicationSet.Do for communicating with the ingesters from the queriers. This mimicks the behavior of Cortex. An extra_query_delay flag has been added to match the flag present in Cortex's querier config. Fixes grafana#1447.
rfratto
added a commit
that referenced
this issue
Jan 6, 2020
This commit uses ReplicationSet.Do for communicating with the ingesters from the queriers. This mimics the behavior of Cortex. An extra_query_delay flag has been added to match the flag present in Cortex's querier config. Fixes #1447.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The Cortex querier, when querying ingesters, only waits for a response from
#ingesters - (replicationFactor / 2)ingesters. Loki is currently waiting for a response from all ingesters.Assuming a replication factor of 3 with 10 ingesters, we only need to wait for a response from 9 ingesters. This optimization can make some queries perform better: occasionally, a single ingester can take significantly longer to return a response.
This should be an easy thing to implement, see Cortex's pkg/ring.ReplicationSet.Do method, which provides the functionality mentioned above.
The text was updated successfully, but these errors were encountered: