Any help would be appreciated.

Are you running with less than 100% BOINC CPU?

Yes, one runs 66% another 75%.
I do have one other PC that R@H/BOINC is running at 60% without exhibiting these problems.

What other projects are you running on these machines?

Seti@home and R@H run on all three machines.

That is why intermittent problems are so hard to find ...

The lockfile problem has been traced to the use of the BOINC "throttling" method. It may also be related to the screensaver and graphics system.

What many have seen is changing to 100% CPU use setting causes these issues to go away. I had to stop RaH for awhile till I ran across a note on EaH about this and tried the 100% setting (I had been at 99%, just to give me a little more "liveliness" in response on that system).

Try it and see if it helps ... if it does not, I will be interested in that too ... but, so far, anyone that has tried it has had the problem go away when they set to 100% CPU use ...

Remember, 100% CPU is often not recommended for laptops and even some desktops due to them being so crowded inside that they'll overheat if run at 100% CPU.

This program is one of those which can help this situation by checking what the current temperatures are:

http://www.almico.com/sfdownload.php

I'm trying to use it - but my motherboard doesn't seem to be on the list of motherboards a set of labels are available for.

On my other machine, though, I've set 100% CPU instead of the 95% CPU I use on this one I use to look for evidence that the lockfile problem has been fixed properly, and don't remember seeing any clear evidence of the lockfile problem on that machine since the change, despite the higher number of workunits it runs.

Also note that some BOINC projects don't allow enough characters for the current setting to allow setting it to 100% without first erasing the .0 they add to the end of the current setting.

Remember, 100% CPU is often not recommended for laptops...

I had that problem due to blocked vents and I was recommended to get one of these: 3-fan Cooler pad

Problem solved.

Several times now on this machine, I've had to kill a mini-rosetta1.67 task because it refused to stop running on demand. BOINC manager thinks all work is suspended due to user activity, yet mini-rosetta 1.67 is using 100% CPU. I hate losing the time wasted, but I can't have Rosetta holding up my normal work.

Is this a known issue? Is there something I need to do to correct this?

Thanks
Tom

---

That is why intermittent problems are so hard to find ...

The lockfile problem has been traced to the use of the BOINC "throttling" method. It may also be related to the screensaver and graphics system.

What many have seen is changing to 100% CPU use setting causes these issues to go away. I had to stop RaH for awhile till I ran across a note on EaH about this and tried the 100% setting (I had been at 99%, just to give me a little more "liveliness" in response on that system).

Try it and see if it helps ... if it does not, I will be interested in that too ... but, so far, anyone that has tried it has had the problem go away when they set to 100% CPU use ...

Hi Paul,

Thanks for your input on this. Strange occurances for sure. Though I do have one machine, out of the three, that has been throttled all along, and it did not exhibit this problem that the other two machines had. These types of inconsistencies sure makes it difficult to troubleshoot the issue.
Like you, I have found I've had to throttle to keep the 'liveliness' up on the machines for when we pesky humans must use these machines.

I'm glad to hear that problem was able to be traced.

It seems that over the last day or so, having updated to BOINC 6.6.36 and now running R@H 1.75, I've not had the same troubles. Work Units are completing and posting, and I'm receving credits on those WUs again. No more 'Compute Error's, Yeah!

Again, thanks for your help on this.