Bug 1349879

+++ This bug was initially created as a clone of Bug #1305031 +++

Description of problem:

AFR winds a few reads of a file in metadata split-brain.

How reproducible:
Always

Steps to Reproduce:
1. cat a big file that is in metadata split-brain
2. It spews out a lot of data before eventually failing it with EIO.



Actual results:
readvs are sent to the brick

Expected results:
No readv should be wound to the brick.

Additional info:

--- Additional comment from Vijay Bellur on 2016-02-05 06:37:25 EST ---

REVIEW: http://review.gluster.org/13389 (afr:Don't wind reads for files in metadata split-brain) posted (#1) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Vijay Bellur on 2016-03-01 08:03:45 EST ---

REVIEW: http://review.gluster.org/13389 (afr:Don't wind reads for files in metadata split-brain) posted (#2) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Vijay Bellur on 2016-03-01 08:07:49 EST ---

REVIEW: http://review.gluster.org/13389 (afr:Don't wind reads for files in metadata split-brain) posted (#3) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Vijay Bellur on 2016-03-01 23:07:13 EST ---

REVIEW: http://review.gluster.org/13389 (afr:Don't wind reads for files in metadata split-brain) posted (#4) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Mike McCune on 2016-03-28 19:24:30 EDT ---

This bug was accidentally moved from POST to MODIFIED via an error in automation, please see mmccune with any questions

--- Additional comment from Vijay Bellur on 2016-06-23 05:20:56 EDT ---

REVIEW: http://review.gluster.org/13389 (afr:Don't wind reads for files in metadata split-brain) posted (#5) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Vijay Bellur on 2016-06-24 07:26:05 EDT ---

COMMIT: http://review.gluster.org/13389 committed in master by Pranith Kumar Karampuri (pkarampu) 
------
commit 7a1c1e2904701496968ed14b6d7479fb706c3188
Author: Ravishankar N <ravishankar>
Date:   Fri Feb 5 15:10:06 2016 +0530

    afr:Don't wind reads for files in metadata split-brain
    
    Problem: For a read on  a file in metadata split-brain:
    1.lookup_done resets event_generation to zero.
    2. readv is issued, goes to inode refresh due to mismatching event_gen.
    3. After refresh is successful, we update event_generation, data and
    metdata readable.
    3. We then call afr_read_txn_refresh_done() which in turn calls
    afr_inode_get_readable() but doesn't check for EIO. So afr_readv_wind
    is called with local->readable (which is populated with data_readable),
    thus winding the read to a brick.
    4. Also, further parallel reads that come directly go to the wind path
    because there is no inode_refresh needed.
    
    Fix:
    1.For any afr_read_txn(), readable must be an intersection of data and metadata
    readable.
    2.Check for EIO in afr_read_txn_refresh_done().
    
    Change-Id: I22dd221fdfaf96d7aced2f474e28ed1337d69f0e
    BUG: 1305031
    Signed-off-by: Ravishankar N <ravishankar>
    Reviewed-on: http://review.gluster.org/13389
    Reviewed-by: Ashish Pandey <aspandey>
    Smoke: Gluster Build System <jenkins.org>
    NetBSD-regression: NetBSD Build System <jenkins.org>
    CentOS-regression: Gluster Build System <jenkins.org>
    Reviewed-by: Pranith Kumar Karampuri <pkarampu>