You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As we know that the slow L0 -> L1 compaction will cause write stall which cause high latency for online application.
Rocksdb have added IOPriority::IO_MID for rate limiter. It's better to all L0-L1 compaction use IO_MID to reduce the writing stall.
Expected behavior
Use can config the Rocksdb to use IO_MID for L0 -> L1 compaction for rate limiter.
Actual behavior
Steps to reproduce the behavior
The text was updated successfully, but these errors were encountered:
As we know that the slow L0 -> L1 compaction will cause write stall which cause high latency for online application.
Rocksdb have added IOPriority::IO_MID for rate limiter. It's better to all L0-L1 compaction use IO_MID to reduce the writing stall.
Expected behavior
Use can config the Rocksdb to use IO_MID for L0 -> L1 compaction for rate limiter.
Actual behavior
Steps to reproduce the behavior
The text was updated successfully, but these errors were encountered: