Details
-
Sub-task
-
Status: Resolved
-
Major
-
Resolution: Duplicate
-
None
-
None
-
None
-
None
Description
Currently, Ratis does a writeStatemachine and readstateMachine data which calls into stateMachine which may actual read actual data from disk. If the disk is overloaded/slow, it can timeout and slow down the whole pipeline. the idea is to add a metric to track these timeouts.
Attachments
Issue Links
- duplicates
-
RATIS-1408 Add metrics of statemachine read/write timeout count
- Resolved