[
https://issues.apache.org/jira/browse/IGNITE-1079?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14903526#comment-14903526
]
Andrey Gura edited comment on IGNITE-1079 at 9/23/15 12:12 PM:
---------------------------------------------------------------
There are two reasons of test failing:
# After killing random node cluster manager chooses the same node as failover.
In this case failover fails but test doesn't provide handler for this
situation. I tested {{Ignite.cluster().nodes}} functionality. It always returns
correct nodes list. Need additional investigation.
# While nodes handle failover after killing random node at the same moment of
time the next node is killing and parallel failover running. The test fails
because test's state changes concurrently.
Debug continues.
was (Author: agura):
There are two reasons of test failing:
# After killing random node cluster manager chooses the same node as failover.
In this case failover fails but test doesn't provide handler for this
situation. I tested {{Ignite.cluster().nodes}} functionality. It always returns
right nodes list. Need additional investigation.
# While nodes handle failover after killing random node at the same moment of
time the next node is killing and parallel failover running. The test fails
because test's state changes concurrently.
Debug continues.
> Implement Ignite-based cluster manager for Vert.x framework.
> ------------------------------------------------------------
>
> Key: IGNITE-1079
> URL: https://issues.apache.org/jira/browse/IGNITE-1079
> Project: Ignite
> Issue Type: New Feature
> Components: cache
> Reporter: Alexey Kuznetsov
> Assignee: Andrey Gura
> Priority: Trivial
> Labels: newbie
> Fix For: ignite-1.4
>
>
> Implement Ignite-based cluster manager for Vert.x framework: http://vertx.io/
> vertx-hazelcast could be used as start point:
> https://github.com/vert-x3/vertx-hazelcast
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
