Description of problem: ======================== After updating one of the two node cluster from rhgs 2.1.6 to 3.1.2, done the volume restart ( Distributed volume ) before updating the other node and updated the other node to 3.1.2 and checked the peer status, peer status was "Rejected" on both the nodes. Version-Release number of selected component (if applicable): ============================================================= glusterfs-3.7.5-6 How reproducible: ================= 100% Steps to Reproduce: =================== 1. Have two node cluster with rhgs 2.1.6 2. Create Distributed and replica volumes using both the nodes. 3. Update one of the node to 3.1.2 // Follow in-service update steps 4. After updating start the glusterd and Distributed volume on the updated node. 5. Now Update the second node to 3.1.2 // Follow in-service update steps 6. After updating second node, start the glusterd and stopped volume (Distributed volume) 7. Check the peer status now on both the nodes Actual results: =============== Peer status showing "Peer Rejected" if we do volume restart before updating the entire cluster Expected results: ================= Peer status should shows connected even if we do volume restart before entire cluster update. Additional info: ================ glusterd log had check-sum mismatch error message. ( Don't have setup to attach it here)
This is being cause due to the change http://review.gluster.org/5568 . This was backported upstream to the release-3.4 branch, which is why we haven't faced any problems like this upstream. This change missed a backport to downstream rhs-2.1 (which is based off release-3.4), but is present in rhs-3.0 and rhgs-3.1, which is causing the problems now. The fix for this is simple enough and only requires change in rhgs-3.1 (and in any future releases which will support upgrade from 2.1). The store function which saves the two additional fields in the volinfo file, needs to save these two lines provided the cluster op-version is greater than or equal to GD_OP_VERSION_RHS_3_0.
Downstream patch: https://code.engineering.redhat.com/gerrit/#/c/61774/
Quota related patch submitted upstream: http://review.gluster.org/12642
As per #comment2 we'd need a downstream fix here which is https://code.engineering.redhat.com/gerrit/#/c/61774/
Verified this bug using the below doc bug[1] info, where volume operations are not supported during the upgrade using glusterfs-3.7.5-12 [1]-https://bugzilla.redhat.com/show_bug.cgi?id=1288049. As per the Document, upgrade is working fine without any issues. Moving this bug to verified state.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHBA-2016-0193.html