[
https://issues.jboss.org/browse/MODCLUSTER-449?page=com.atlassian.jira.pl...
]
Michal Babacek commented on MODCLUSTER-449:
-------------------------------------------
Wouldn't it be better addressed in the native part? The problem is not the load number
itself, it's the {{elected}} counter. I suggest treating "newcomers" gently
in {{internal_find_best_byrequests}}. IMHO one could guess that info from a large
discrepancy between the elected counters of other nodes and this new one. I'm gonna
try to verify I'm not just being full of ideas...
Implement ramp-up when starting new nodes
-----------------------------------------
Key: MODCLUSTER-449
URL:
https://issues.jboss.org/browse/MODCLUSTER-449
Project: mod_cluster
Issue Type: Feature Request
Components: Core & Container Integration (Java)
Reporter: Radoslav Husar
Assignee: Radoslav Husar
Priority: Critical
Fix For: 1.3.1.Beta3
IIUC this has been a problem since inception. The problem is that the initial load stays
in effect for performing load-balancing decisions until a new stat interval kicks in.
This effect is mitigated by load decay over time, but for the time a new node joins in,
it can get overloaded upon startup.
--
This message was sent by Atlassian JIRA
(v6.3.11#6341)