[
https://issues.apache.org/jira/browse/IGNITE-7786?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dmitry Sherstobitov updated IGNITE-7786:
----------------------------------------
Description:
Looks like there is hardcoded timeout for waiting result of change baseline
operation
In big cluster there is following behaviour: (154 nodes)
# Set new baseline topology version
# Utility connects, but then fails by connection error
# Cluster successfully activated
{code:java}
...Start node...
...Waiting for topology snapshot...
> control_utility.sh --baseline version 9
Control utility
2017 Copyright(C) Apache Software Foundation
User: test
--------------------------------------------------------------------------------
Failed to set baseline with specified topology version.
Connection to cluster failed.
Error: Failed to perform request (connection failed): /IP
...few milliseconds later...
> control_utility.sh --baseline version 9
Control utility
2017 Copyright(C) Apache Software Foundation
User: test
--------------------------------------------------------------------------------
Cluster state: active
Current topology version: 9
Baseline nodes:
ConsistentID=node1, STATE=ONLINE
ConsistentID=node10001, STATE=ONLINE
ConsistentID=node2, STATE=ONLINE
ConsistentID=node3, STATE=ONLINE
ConsistentID=node4, STATE=ONLINE
--------------------------------------------------------------------------------
Number of baseline nodes: 5
Other nodes not found.{code}
was:
Looks like there is hardcoded timeout for waiting result of change baseline
operation
In big cluster there is following behaviour: (154 nodes)
# Set new baseline topology version
# Utility connects, but then fails by connection error
# Cluster successfully activated
> Changing baseline topology on cluster may have error in control.sh utility
> --------------------------------------------------------------------------
>
> Key: IGNITE-7786
> URL: https://issues.apache.org/jira/browse/IGNITE-7786
> Project: Ignite
> Issue Type: Bug
> Affects Versions: 2.3
> Reporter: Dmitry Sherstobitov
> Priority: Major
>
> Looks like there is hardcoded timeout for waiting result of change baseline
> operation
> In big cluster there is following behaviour: (154 nodes)
> # Set new baseline topology version
> # Utility connects, but then fails by connection error
> # Cluster successfully activated
> {code:java}
> ...Start node...
> ...Waiting for topology snapshot...
> > control_utility.sh --baseline version 9
> Control utility
> 2017 Copyright(C) Apache Software Foundation
> User: test
> --------------------------------------------------------------------------------
> Failed to set baseline with specified topology version.
> Connection to cluster failed.
> Error: Failed to perform request (connection failed): /IP
> ...few milliseconds later...
> > control_utility.sh --baseline version 9
> Control utility
> 2017 Copyright(C) Apache Software Foundation
> User: test
> --------------------------------------------------------------------------------
> Cluster state: active
> Current topology version: 9
> Baseline nodes:
> ConsistentID=node1, STATE=ONLINE
> ConsistentID=node10001, STATE=ONLINE
> ConsistentID=node2, STATE=ONLINE
> ConsistentID=node3, STATE=ONLINE
> ConsistentID=node4, STATE=ONLINE
> --------------------------------------------------------------------------------
> Number of baseline nodes: 5
> Other nodes not found.{code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
