Details
-
Improvement
-
Status: Closed
-
Major
-
Resolution: Invalid
-
None
-
None
-
None
-
None
-
HBase Mgmt User Interface
Description
It would be great if the HBase management user interface provided more bullet-proof reporting in terms of region counts and status. The regions counts go up and down and different views show different counts and they don't always add up. It makes one less than confident of what is going on.
Cluster wide And per table it would be great to get summary counts of all regions in the various states. It would also be useful to see this state in the description of the region. There is a compaction count, but what about in split, memstore flush, read/write lock, disabled, etc. Fundamentally this is the health monitor of the system and as a dba one really needs to know the 100% count of regions and where they are all at in terms of availability. Are they disabled, blocked for writes, blocked for reads, in compaction, etc. etc.
The first step is to define and document exactly what states a region could be in and what impact that state could have on performance, write blocking, read blocking, compaction, disabling, etc. etc. This needs to be defined and well documented. Once these states are defined and well understood (and explain every problem one could have reading/writing/pausing/disabling) then adding to the UI view to show a count of all regions within each state (with a percentage) and also report this state in the detailed list of regions it would give one an holistic view of exactly what is going on.