Bug 1349792

+++ This bug was initially created as a clone of Bug #958419 +++

Description of problem:
gluster does not show a true volume status field. The status of the volume should account for the state of the bricks that make up the volume, and also activities that impact the volume like brick removal and addition.

This BZ is therefore an RFE to change the way the volume state is derived in the CLI.

Version-Release number of selected component (if applicable):
2.0 update 4 onwards

How reproducible:
RFE request

Steps to Reproduce:
1. vol info -  to get a brick list
2. vol status - to show the status of the bricks
3. take a node down and repeat the volume status command
  
Actual results:
At the moment, vol info shows the bricks that should be available to the volume. volume status shows those bricks - but if a node is down, the status doesn't show the missing brick at all (see BZ886865) - and more importantly the volume's state does not reflect the fact that the volume is effectively degraded (lost resiliency).

Expected results:
The volume status field should account for the components of the volume and any lifecycle operations that may be affecting the volume. 

For example, in a dist-repl volume 
- when all bricks are online - status = healthy
- when one brick/node in a replica set is down, status is "degraded"
- when multiple bricks are offline, but the offline bricks still leave an alternate brick within the replica set available - status remains "degraded"
- when a 2 bricks in the same replica set go offline - status should show something like "restricted" to indicate that data access will be impaired.
- when all bricks are inaccessible for a volume - status should show "Offline"

for a distributed volume it's a little simpler - it's either healthy, or restricted

In addition, when bricks are added to a volume the rebalance process should trigger a volume state change to 'Growing', and when removing bricks - "Shrinking"..

Additional info:

--- Additional comment from RHEL Product and Program Management on 2013-05-01 08:24:32 EDT ---

Since this issue was entered in bugzilla, the release flag has been
set to ? to ensure that it is properly evaluated for this release.

--- Additional comment from Simon Green on 2013-07-28 19:18:42 EDT ---

User pcuzner's account has been closed

--- Additional comment from John Skeoch on 2015-11-03 18:07:52 EST ---

User kparthas's account has been closed

--- Additional comment from Atin Mukherjee on 2015-12-24 04:13:37 EST ---

Agreed Paul, this is a limitation in volume status implementation. We have plans to fix this in GlusterD 2.0 as we can rely on centralized store, volume status can still be a local glusterd operation with out talking to all other nodes. We plan to re target this bug for RHGS 4.0. 

I am still keeping this bug open till we have rhgs 4.0 flag available.

--- Additional comment from Atin Mukherjee on 2016-06-24 05:36:16 EDT ---

I am going to clone this bug under glusterd2 component and close this now.