"Tried to execute unknown prepared query" error after a node restart

Description

I use a prepared statements cache in my application (so, I prepare a query only once before the first usage).
The cluster configuration is: 5 nodes, RF = 3, CL = LOCAL_QUORUM for read and write.
After node restart I got a lot of "Tried to execute unknown prepared query" errors, which stops immediatly after the application restart.
As I can understand I could invalidate my applications cache after a node joining in cluster, but it seems like ugly hack.

Environment

datastax driver 2.0.0-rc3, Cassandra 2.0.5 (5 nodes in cluster)

Pull Requests

None

Activity

Show:
Vladimir Kuptsov
March 25, 2014, 7:19 PM
Edited

I've checked the 2.0.1 driver version and it works fine in the case described early.

Sylvain Lebresne
February 28, 2014, 7:36 PM

I'll refer you to since that's rather likely a duplicate of it (it's at least the same error, so even if it's not the same cause, I'd rather focus on fixing all instance of the problem on ). I did pushed some fix related to it in the 2.0 final, but haven't closed the ticket yet since no-one running into the issue has yet confirmed that this did fixed the issue. So can you give a shot to the 2.0 final and see if it helps? And if it doesn't, I'd still be happy to know on (and in that case, if you have a relatively easy way to reproduce, that would be very helpful).

Duplicate

Assignee

Unassigned

Reporter

Vladimir Kuptsov

Affects versions