We are using grinder-3.1.1 on linux. (VMs, 4 of them), with Python/Jython.
vm1: console and agent1
We have seen from time to time, our jython script changes not making it to the agents, even after we force distribute them from the console. It usually "takes" in two tries. We sometimes have to kill the agents and restart them too.
Other teams that use our grinder setup (it is shared) see this too.
All teams are very technical people, and report the same behaviour.
Any idea why a distribute wouldn't take?
My co-worker the grinder guru says
problem here is grinder is not able to update file store on first
attempt. It tries to update the file store but when you do execution, it
uses the older version of file store"