| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: Beta Testing Help Conquer Cancer Linux |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 15
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I've received one of these. Does "This is intended to reduce the number of errors that occur in the app on Linux" mean it's supposed to get rid of the SEGV at the end, or just the other problems?
Sadly, the machine which received it can't possibly run it in the time allocated, so I'll have to abort it anyhow, but it'd be interesting to know. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi Kremmen,
A few weeks ago, armstrdj was working on an interaction between some versions of the libc library and the pthreads library which he believed was causing the crashes in Linux. You will remember that he asked you which version of glibc you were running. My guess is he has bundled both the crash fix and the memory management changes together. Just a guess on my part. Now that I have put my money down, we will see if I win or lose.  Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Not sure how soon we'll find out. The scheduler keeps giving a slow machine betas which would take it days with a short deadline. It keeps giving them to that machine (eg. BETA_ X0000045461373200502081407_ 4-- with deadline under 10 hours) and not to faster machines. I would have thought all the aborts would have disqualified the slow machine from getting beta jobs, even if the scheduler somehow didn't notice that it can't finish them, but it keeps giving them to the slow, unreliable machine instead of the faster, reliable machine.
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Select the profiles with which you want to run Beta (e.g. the fastest or the machine that had a specific project problem) and attach the relevant device to that profile.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Also, who cares about deadlines on Beta units. That is just a way to tell BOINC Manager to do a fast turn-around on the Beta unit.
|
||
|
|
armstrdj
Former World Community Grid Tech Joined: Oct 21, 2004 Post Count: 695 Status: Offline Project Badges: 
|
Kremmen,
Yes this beta test includes a potential fix for the SEGV issue towards the end of the run. If the slow machine that is running the beta workunit has had this error in the past I recommend letting it run to completion to see if it now exits successfully. If one of your faster machines that experiences the error gets a beta workunit as well then you can abort the other one if you like. Thanks, armstrdj |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Also, who cares about deadlines on Beta units. That is just a way to tell BOINC Manager to do a fast turn-around on the Beta unit. Ordinary units which run over time in some way get aborted by the WCG server, so I assumed beta ones would too. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
A less optimistic piece of news on the Beta for HCC on Linux posted in the back room.... sit tight as the programmers plod along:
----------------------------------------The HCC1 beta test did not show improvement for the Linux sig 11 problems. So we are looking at that further.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A less optimistic piece of news on the Beta for HCC on Linux posted in the back room.... sit tight as the programmers plod along: Thanks for letting us know. I wonder if they'll get it fixed by the time AC@H ends? (But then I'm still wondering if they'll ever just isolate the little bit of faulty code so that the problem can be tested 4 orders of magnitude faster than by running a whole WU.) |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Any news on any progress on fixing this yet?
----------------------------------------Edit: Now holding breath to see how the version 6.02 beta goes. [Edit 1 times, last edit by Former Member at Apr 10, 2008 1:53:08 PM] |
||
|
|
|