Description
As discussed on HDFS-6682, admins currently have very little introspection into how fast recovery work is progressing on the cluster. It'd be useful to have rate metrics for the pending queue on the NN, possibly also DN-side too.
Attachments
Issue Links
- is related to
-
HDFS-6682 Add a metric to expose the timestamp of the oldest under-replicated block
-
- Resolved
-