-
Type:
Task
-
Status: Resolved
-
Priority:
Minor
-
Resolution: Fixed
-
Affects Version/s: None
-
Fix Version/s: None
-
Component/s: Performance
-
Epic Link:
ES configuration
The AWS recommendation is to limit shards size to 45GB.
based on the current metrics, if we want to go to 1B on the current configuration it means that we need to target 35 shards.
For the replication, we clearly do not need 2 of them: just one should be enough.
ES deployment
We are using AWS Elasticsearch service, the recommendation is to isolate data nodes from master nodes:
- 3 data nodes
- 3 master nodes
Stream configuration
We want to apply the perf template config.
We also want to increase the concurrency for reIndex.
