I'm running on Amazon's EC2 and "sometimes" find myself with ephemeral
nodes after all of my workers have exited or been terminated.
I've seen other reports of this problem that turned out to be user error
after one of the Zookeeper maintainers dumped the transaction log and found
a problem. (See Jira issue ZOOKEEPER-1600)
I'd love to solve this problem myself but the transaction log looks binary
to me. I see references to my nodes and paths but any actions look binary.
Is there a way to decipher the log?
Camille Fournier 2013-03-25, 19:59
Brian Tarbox 2013-03-25, 20:28