mxm commented on PR #13382: URL: https://github.com/apache/iceberg/pull/13382#issuecomment-3018671496
You are right Peter, the cache size should probably be proportional to the "active" tables, but using a time-based eviction policy, we have seen worse performance. That was the reason we switched to max size / LRU. So there should probably be an LRU eviction policy not across ALL tables, but per-table. We could extend LRUCache to use a LinkedList per table. We probably have to then compare it again to something like Caffeine to decide if the custom cache implementation is still worth it. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
