ClickHouse / ClickHouse/ClickHouse
Does ClickHouse Need a Remote Query Cache?
- Dominant language
- C++
- Stars
- 49.9k
- Forks
- 9k
- Avg merge
- 21h 32m
- Merged PRs (30d)
- 515
Description
### Company or project name
-
### Use case
-
### Describe the solution you'd like
I am a heavy user of ClickHouse, managing a cluster with over 5,000 nodes. Currently, I am facing an issue where different nodes in the cluster receive a large number of duplicate queries in a short period. When these repeated queries have long execution times, the cluster's load increases significantly, and OOM (Out of Memory) errors become more likely. The existing QueryCache does not solve this problem because the cluster has many compute nodes, and identical queries are often distributed across different nodes, making cache hits impossible.
To address this scenario, I am considering implementing a RemoteQueryCache that stores query results in an external system like Redis, enabling query result sharing across compute nodes. Would this solution be valuable? If implemented, would it have a chance of being merged into the ClickHouse?
### Describe alternatives you've considered
-
### Additional context
-
Contributor guide
Assessment
This issue has not been assessed yet.