The 1.0 branch doesn't work with cassandra-1.0.10 about priam HOT 5 CLOSED

Hmmm, I haven't seen this before. We are running the latest build(s) of the 1.0 branch internally here and haven't had problems like the one you are seeing. Also, we're starting to the trunk version for cassandra 1. (in test), as well.

Is this the only exception that gets thrown, or are there others? Also, are you adding in any extra jars/config to the webapp?

Also, maybe setting the tomcat debugging level really low may help print out some extra info we can use to help debug,

from priam.

It's been an active week, and my half-broken priam-1.1 magically recovered 2 nodes that died on Saturday - Hooray!

I'm pretty ignorant of java, so I'm 100% sure that I just ran mvn with the recommended flags from the docs. Would the version of Tomcat I'm using matter?

from priam.

We're using Tomcat 6.x here, and that has been stable for a long time with Priam. A google search yielded this link :http://stackoverflow.com/questions/6564553/tomcat-6-log4j-1-2-16-java-lang-illegalstateexception-class-invariant-viol . How are you deploying the Priam webapp to Tomcat? Do you bounce tomcat each time, or just redeploy the war to a running Tomcat process?

from priam.

I've tried clearing the property referenced in stackoverflow - it gets rid of the invariant error message. In my case that message happens when the server has detected an error and has already unloaded it, so I'm still without priam in this case.

My deployment, when testing, is to shut down tomcat6, delete all the state I find in /var/lib/tomcat6/webapp/Priam* and one or two other places I can't recall right now, then replace with a new war and start up tomcat again. When it's died once this seems to be minimum to get a working system.

from priam.

It turns out that updating dependencies in maven has fixed this. I assume the gradle script in the current master branch would do the same.

Closing.

from priam.