[
https://issues.jboss.org/browse/JGRP-1910?page=com.atlassian.jira.plugin....
]
Matej Čimbora commented on JGRP-1910:
-------------------------------------
While testing this on ISPN I found out the issue still occurs from time to time, i.e. node
receives a merge view, which is just a subset of the previous one. The attached test
passes, however when we set non-null discard rate, the issue is reproducible (from logs),
e.g.
[Incoming-1,MergeStressTest,3] accepted MergeView::[3|10] (7) [3, 4, 8, 9, 1, 0, 2], 3
subgroups: [3|8] (2) [3, 4], [8|9] (2) [8, 9], [1|8] (3) [1, 0, 2]
[Incoming-2,MergeStressTest,3] accepted MergeView::[3|11] (4) [3, 4, 8, 9], 2 subgroups:
[3|8] (2) [3, 4], [8|9] (2) [8, 9]
I'm attaching an updated version of the test which fails as soon as the issue is
detected. Can you please confirm this behavior is unexpected so that further
investigations can be performed?
MERGE3: Do not lose any members from view during a series of merges
-------------------------------------------------------------------
Key: JGRP-1910
URL:
https://issues.jboss.org/browse/JGRP-1910
Project: JGroups
Issue Type: Bug
Reporter: Radim Vansa
Assignee: Bela Ban
Fix For: 3.6.3
Attachments: SplitMergeTest.java
When connection between nodes is re-established, MERGE3 should merge the cluster
together. This often does not involve a single MergeView but a series of such events. The
problematic property of this protocol is that some of those views can lack certain
members, though these are reachable.
This causes problem in Infinispan since the cache cannot be fully rebalanced before
another merge arrives, and all owners of certain segment can be gradually removed (and
added again) to the view, while this is not detected as partition but crashed nodes ->
losing all owners means data loss.
Removing members from view should be the role of FDx protocols, not MERGEx.
--
This message was sent by Atlassian JIRA
(v6.3.11#6341)