[
https://issues.jboss.org/browse/ISPN-10821?page=com.atlassian.jira.plugin...
]
Tristan Tarrant updated ISPN-10821:
-----------------------------------
Sprint: DataGrid Sprint #35, DataGrid Sprint #36, DataGrid Sprint #37 (was: DataGrid
Sprint #35, DataGrid Sprint #36)
Decrease GMS join_timeout and increase xPING num_discovery_runs
---------------------------------------------------------------
Key: ISPN-10821
URL:
https://issues.jboss.org/browse/ISPN-10821
Project: Infinispan
Issue Type: Task
Components: Configuration
Affects Versions: 9.4.16.Final, 10.0.0.CR3
Reporter: dan.berindei
Assignee: dan.berindei
Priority: Major
Fix For: 10.0.0.Final
Our default JGroups configurations have a GMS {{join_timeout="5000"}}, meaning
the first node to start will always wait for 5 seconds before becoming coordinator. It
should be safe to use the JGroups default of 2 seconds.
To make it even more safe, we could configure the discovery protocol's
{{num_discovery_runs="3"}}, so that it makes multiple attempts to contact
existing nodes during those 2 seconds of waiting. {{num_discovery_runs}} is only used for
the initial discovery, so there's no danger of creating additional traffic while the
node is up.
--
This message was sent by Atlassian Jira
(v7.13.8#713008)