So the big bug that happened after the server code was updated was that there wasn't a working limit for GPU work, which means a lot of clients downloaded thousands of workunits, crippling the database; and even worse crippling the disk of the server even harder reporting all of those workunits. I'm pretty sure the only way we're going to be able to get the server back up and running in any reasonable amount of time is to clear out the entire workunit and result tables. They're so large I can't even query their size. So basically to get things up and running again I'll be clearing out the result and workunit tables of the database -- we're hoping we can get that done sometime this weekend. As well as cleaning out our upload and download directories, which are slow almost to the point of unresponsiveness. After things are cleared out, we'll bring things back up (with working workunit limits for both CPUs and GPUs) and I'm hoping things will run fairly smoothly after that. I know it sucks and I wish we had a better solution, but I think this is the best we can do. Feel free to cancel any workunits/results you have currently crunching; and try not to flame us too hard. :P
More...
Reply With Quote
