Github user serg38 commented on the issue:
https://github.com/apache/cloudstack/pull/1762
@rafaelweingartner Thanks a lot. I totally agree that resetting hosts
doesn't really need to be a part of transaction and should be extracted to a
new method. The same is for lines 527-546, and then another one after 551
My understanding of agent LB is that is handled separately from reconnect
part. I might be wrong but it is done in ClusteredAgentManagerImpl by
scheduling rebalancing task every 60 sec
getAgentRebalanceScanTask which takes care of transferring of connected
agents.
@rhtyd @jburwell @koushik-das @karuturi Do you agree that we can split a
transaction in findAndUpdateDirectAgentToLoad into 3 non transactional methods
and thus eliminate a one side of a repeated deadlock? This is a very core of
agent management that is very hard if ever possible to write smoke test. If so
@nvazquez might be able to work on refactoring this method later this month
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
