Can someone tell me why these WU's errored out.
4131418
4130849
4124171
4122129
4119432
Is it on my end or faulty WUs? I have processed 1 good one out of 6.

---

Can someone tell me why these WU's errored out.
4131418
4130849
4124171
4122129
4119432
Is it on my end or faulty WUs? I have processed 1 good one out of 6.

Neither, it's a problem with the boincapi when Cuda22 tasks are suspended or pre-empted, a fix has been put in place that will be in the next app,

Claggy

We think there's a problem with the cuda22 application (only) - it doesn't like being interrupted in the middle of a task. Unfortunately, cuda22 is the only version which will run on your 256 MB card - all the other versions need more memory.

If you can leave them to run for long enough to complete the task in a single run, you should be OK. Your successful tasks - two of them now - ran for around 3 hours, so that may need some care: if they are being paused so that another project can take a turn, you can avoid that by increasing the 'Switch between tasks every' value in Computing preferences. There's not much you can do to avoid it if the tasks are still running when you need to turn the computer off, unfortunately.

There is a fixed version of the cuda22 app now, but we are still investigating possible problems (completely unrelated) with the cuda32 app on some cards. I would expect that both fixes will be deployed together for the next phase of testing, so it may be a few days before these run properly - but it will be very useful to have your input then, as there are not many active testers with these smaller cards.

We think there's a problem with the cuda22 application (only) - it doesn't like being interrupted in the middle of a task. Unfortunately, cuda22 is the only version which will run on your 256 MB card - all the other versions need more memory.

If you can leave them to run for long enough to complete the task in a single run, you should be OK. Your successful tasks - two of them now - ran for around 3 hours, so that may need some care: if they are being paused so that another project can take a turn, you can avoid that by increasing the 'Switch between tasks every' value in Computing preferences. There's not much you can do to avoid it if the tasks are still running when you need to turn the computer off, unfortunately.

There is a fixed version of the cuda22 app now, but we are still investigating possible problems (completely unrelated) with the cuda32 app on some cards. I would expect that both fixes will be deployed together for the next phase of testing, so it may be a few days before these run properly - but it will be very useful to have your input then, as there are not many active testers with these smaller cards.

I do run several cuda projects and switch between them every 120min. I will set the SETI Beta to switch every 480 minutes and see if this helps. I know my card sux (256mb ram) :( I hope to upgrade soon. I haven't been crunching in a while but have started back. I missed the fun ;P! I kinda figured that was why I was having trouble. Thanks for the heads up.

We think there's a problem with the cuda22 application (only) - it doesn't like being interrupted in the middle of a task. Unfortunately, cuda22 is the only version which will run on your 256 MB card - all the other versions need more memory.

Oh, so that's what was going on. I wondered if there was something wrong with the cuda22 WU's I was running. I was micro managing my cache, suspending some cuda22, to run a few 23 and 32 WU's, and everytime a suspended cuda22 restarted, it always errored out. I noticed though, that if I let them run uninterrupted from start to finish, they would always end normally.

I never thought it was a problem with the cuda22 app itself. Thanks for explaining, and I hope a new version is released soon.

Regarding Cuda22 erroring out. This one: resultid=11297800 errored out after about 30 minutes, and it was never suspended or restarted. It run from start and only 30 minutes, before it crashed and burned.

So it seems as if Cuda22 can error out randomly too, and not only if they are interrupted and restarted.