How to end a work unit.
- Thread starter muon
- Start date
Most WU's take a while, I just plod through them. 3402's and the like are a bit soul destroying though.
To start and stop the client if you have them installed as a sevice, I always use services.msc.
There is a command to terminate a WU early but that's probably frowned on.
To start and stop the client if you have them installed as a sevice, I always use services.msc.
There is a command to terminate a WU early but that's probably frowned on.
Last edited:
ghost101 said:
It's not really advisable though, because that WU now has to be reassigned by Stanford which will take months.
There's no way any WU should give 30ppd on an E6300. Possibly it was running without SSE instructions because it wasn't shut down properly? Check the log file for a line like 'Extra SSE boost OK', or 'Working with standard loops on this execution.'
Agree with Mattus. It is really bad for the science to nuke a WU for the hell of it. It should only be done if you have a legitimate reason - ie it is causing errors and you keep getting it or you absolutely know you aren't going to finish [only really applicable to very, very short deadlines].
I would check as above that the client isn't running Standard Loops. 30ppw is microscopic. I'd be surpised if it was 30ppd tbh.
I would check as above that the client isn't running Standard Loops. 30ppw is microscopic. I'd be surpised if it was 30ppd tbh.
Not using the -forceasm flag can make a major difference to the processing speed. I missed it out on one of my borgs and the PPD doubled when I corrected it.
An E6300 should give at least 120 PPD on a 3402 WU so you definitely have a problem, probably configuration.
An E6300 should give at least 120 PPD on a 3402 WU so you definitely have a problem, probably configuration.