Excuse me for starting this thread, but I looked in some of the earlier thread about this problem, where the Rosetta WU's crashed if not staying in memory while preempted, but I couldn't find it, and the threads I looked in, gave no answer to my question.

So, now after upgrading to 5.2.2. is it still necessary to let the WU's stay in memory while preempted??? I don't remember if that problem with the crashing were fixed, or it only was the problem about crashing while running benchmarks, that were fixed?!

---

[b]"I'm trying to maintain a shred of dignity in this world." - Me[/b]

Excuse me for starting this thread, but I looked in some of the earlier thread about this problem, where the Rosetta WU's crashed if not staying in memory while preempted, but I couldn't find it, and the threads I looked in, gave no answer to my question.

So, now after upgrading to 5.2.2. is it still necessary to let the WU's stay in memory while preempted??? I don't remember if that problem with the crashing were fixed, or it only was the problem about crashing while running benchmarks, that were fixed?!

I think it has a possibility of crashing (about 10%) when removed from memory, for whatever reason.

---

Hi. Prob best to wait for one of the staff to answer this question, I would like to know the answer too.
(your kitty looks just like ours}.

---

It looks like it is still happening even using the updated client. Are people still seeing this?

Well, I just joined this project on 20 October 2005 and I've had four errors, out of nine work units. The jury's still out on the two units in progress. I'm running Boinc 5.2.2. I've had two errors and one abort with the 4.77 application and one error with the 4.78 application. The abort was on a unit that hung at 83.3 percent complete, no matter how many times I stopped the core client and application. I've tried warm boots, cold boots and nothing would get the stuck unit to take off. The other three errors appeared to be pauses, in the Boinc manager, to switch to Einstein, when the fatal errors occurred in Rosetta. I was not keeping the application in memory during switches.

There's nothing special about the two boxes I'm running: 1.7Ghz & 1.8Ghz P4's running XP Home edition, SP2 with all current updates. They run Einstein & SETI just fine. I think this is a worthwhile project, and I will gladly donate my computer's idle time, but I would like to see a little better MTBF on the work units.

Steve

Last 120 WUs have no 'exits early'. Running 5.2.2 for a few days now.
IBM Think Centre. Windows SP2, 256MB Ram. 2.8GB, and 750MB in Virtual RAM. Have not noticed any "hanging WUs", at least not to excess.

---

Ok, I won't change my settings then.

I just didn't know if that problem was fixed, but it apparently isn't.

And thanks! :-)

---

[b]"I'm trying to maintain a shred of dignity in this world." - Me[/b]