This last fortnight the cluster has been on and off. There were a few warm days where it was off but for the most part its been running 24/7.
The 12 Pi3B+ nodes have been running Einstein BRP4 work and the 4 Pi3B nodes have been doing a mix of both Seti and Einstein. They are currently back on Seti.
Einstein@home added a new beta app which confused their scheduler so all the Pis were requesting work and getting told there wasn't any - Even though the server status page said there were 10,000 work units available. I hadn't changed anything so I assumed it was a project issue. I set the Pi3B+ nodes to No New Work overnight and allowed the Pi3B's to do Seti work. The following day I enabled work fetch on the Pi3B+ nodes and they all managed to pick up work.
No comments:
Post a Comment