Hi Guillaume,
I wonder what happens if the recovery log db is unaccessible for a
long time.
What I'm expecting for is:
Failure 1
...
Failure n-limit
-> Controller failure
(logging and behavior)
Emmanuel, did I miss something?
No I think this is correct. Long recovery log outages will eventually
lead to an out of memory in the controller that will finally fail.
There is no admission control on the logger thread queue. This could be
filed as an improvement in a JIRA issue.
Thanks for pointing this out,
Emmanuel
--
Emmanuel Cecchet
Chief Scientific Officer, Continuent
Blog: http://emanux.blogspot.com/
Open source: http://www.continuent.org
Corporate: http://www.continuent.com
Skype: emmanuel_cecchet
Cell: +33 687 342 685
_______________________________________________
Sequoia mailing list
[email protected]
https://forge.continuent.org/mailman/listinfo/sequoia
