Message boards : Number crunching : Problems and Technical Issues with Rosetta@home
Previous · 1 . . . 287 · 288 · 289 · 290 · 291 · 292 · 293 . . . 295 · Next
Author | Message |
---|---|
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
Tasks with rb_08_16 and 17 are running good. 2:46 into a 8hr run and no problems so far. and another at 1:50/8 hrs is also running good. My hal_8a_p_hal.......... task crashed immediately, but that looks like old work. |
Grant (SSSF) Send message Joined: 28 Mar 20 Posts: 1631 Credit: 16,677,279 RAC: 8,009 |
I see there are a bunch of Beta 6.06 Tasks Ready to send as well as In Progress. However, the Average computing for those applications is still showing as 0 GigaFLOPS. I'm thinking the Tasks, application or support files for those Tasks are still borked. And the problem with that is when a Task errors out, it results in a delay being added to the next time the manager will contact the Scheudler. With every Task erroring out, you end up with multi-hour delays- the more cores/threads, the more Tasks that error out & the longer the delay (this is by design so that systems producing lots of errors don't end up doing a DoS (Denial of Service) attack on the Scheduler. The logic being that projects won't use their main project for alpha or beta testing their applications...). This also stops those systems from getting work that they could actually process OK until they can contact the Scheduler (unless the user iis prepared to sit there & hit update till all of the duds are cleared out). End result- it is going to take a long time for all of these Tasks to finally clear from the system if the Project doesn't just pull them all now. Grant Darwin NT |
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
I see there are a bunch of Beta 6.06 Tasks Ready to send as well as In Progress. However, the Average computing for those applications is still showing as 0 GigaFLOPS. Depends on which beta you got..as I said..the hal_8 stuff is buggy. But I only got one. So if you have hal_8 take a hit and kill them. the rb_08_16 and 17 is clean. I am running those now. I am halfway through 2 of them with no problem. This is the buggy stuff: https://boinc.bakerlab.org/rosetta/workunit.php?wuid=1407403018 |
Grant (SSSF) Send message Joined: 28 Mar 20 Posts: 1631 Credit: 16,677,279 RAC: 8,009 |
Depends on which beta you got.No, it doesn't- all of the Tasks for the Beta 6.06 application error out. The Rosetta 4.20 application Tasks are OK (other than the usual odd error). They are 2 different sets of Tasks being processed by 2 different applications- Beta 6.06 v Rosetta 4.20. Grant Darwin NT |
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
Depends on which beta you got.No, it doesn't- all of the Tasks for the Beta 6.06 application error out. The Rosetta 4.20 application Tasks are OK (other than the usual odd error). True, didn't notice that in the middle of the night. Rosetta 29845 176784 7.01 (0.15 - 43.54) 4933 Rosetta Beta 18000 6699 ---- 0 <-- being withheld or because the errors are recirculating? |
Richard James Send message Joined: 30 Mar 20 Posts: 14 Credit: 1,998,675 RAC: 1,599 |
>All rosetta tasks so far today have failed within a few seconds or minutes with "computational error" and "output file... absent" in the log. Al running OK now. I see those were "beta" tasks. However, I do not see a way to prevent downloading them, is there? Richard |
[VENETO] boboviz Send message Joined: 1 Dec 05 Posts: 1978 Credit: 9,194,012 RAC: 3,787 |
I see those were "beta" tasks. However, I do not see a way to prevent downloading them, is there? Not from your user profile in this site. You can use, if you want, a configuration file in boinc manager... |
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
I see those were "beta" tasks. However, I do not see a way to prevent downloading them, is there? Richard - your just unlucky right now. Your getting buggy tasks that have to run through two computers to be flagged as buggy. Your just the lucky wingman in this case. Just let the server do its thing, eventually you will get clean work. You can ask Veneto how to do that modification to block beta work. But I only got one beta and since then nothing but clean 4.20. So is it worth the effort to mess around? I don't think so. The beta will soon be done. It looks like nothing new is being sent out. 12,680 tasks was where beta was last night at this time. So you should get clean 4.20 soon. |
kotenok2000 Send message Joined: 22 Feb 11 Posts: 257 Credit: 476,390 RAC: 937 |
Old news are threads are still not readable. https://boinc.bakerlab.org/rosetta/forum_forum.php?id=202&sort=5&start=150 |
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
Old news are threads are still not readable. You have to click the page # below and you can read them...or at least I can. This link you posted goes to page 4. The link direct does not work, but if you manually go to page 4 you can see whats going on. |
[VENETO] boboviz Send message Joined: 1 Dec 05 Posts: 1978 Credit: 9,194,012 RAC: 3,787 |
You can ask Veneto how to do that modification to block beta work. But I only got one beta and since then nothing but clean 4.20. So is it worth the effort to mess around? The "usual" app_config.xml Something like this (if i remember correctly): <app_config> <app> <name>rosetta</name> <max_concurrent>X</max_concurrent> </app> </app_config> (with X you can configure how many cores to use) P.S. Veneto is a region (where i live) of Italy. My nickname is boboviz... |
kotenok2000 Send message Joined: 22 Feb 11 Posts: 257 Credit: 476,390 RAC: 937 |
Stuprd boinc thinks <max_concurrent>0</max_concurrent> is unlimited. I can read news from page 1, but when i press on page 2 everything is empty. |
Grant (SSSF) Send message Joined: 28 Mar 20 Posts: 1631 Credit: 16,677,279 RAC: 8,009 |
Stuprd boinc thinks <max_concurrent>0</max_concurrent> is unlimited.I only get threads on page 1, all the rest are empty. Grant Darwin NT |
Grant (SSSF) Send message Joined: 28 Mar 20 Posts: 1631 Credit: 16,677,279 RAC: 8,009 |
And after making that above post, "Problems and Technical Issues with Rosetta@home" then went missing from Number Crunching for me.Stuprd boinc thinks <max_concurrent>0</max_concurrent> is unlimited.I only get threads on page 1, all the rest are empty. ??? Grant Darwin NT |
kotenok2000 Send message Joined: 22 Feb 11 Posts: 257 Credit: 476,390 RAC: 937 |
I can see it as thread 7. |
Grant (SSSF) Send message Joined: 28 Mar 20 Posts: 1631 Credit: 16,677,279 RAC: 8,009 |
I can see it as thread 7.UPI Payment spam is the most recent post showing for me, from 2 hours ago. But on the main forum page, it shows the most recent post in the Number crunching thread was 2 minutes ago. I've set the messages boards to ignore pinned messages. I can only find this thread now by looking at my recent posts in my account. Grant Darwin NT |
kotenok2000 Send message Joined: 22 Feb 11 Posts: 257 Credit: 476,390 RAC: 937 |
I've set the messages boards to ignore pinned messages. Here is your problem. |
Sid Celery Send message Joined: 11 Feb 08 Posts: 2072 Credit: 40,598,509 RAC: 5,584 |
Some more Robetta tasks have popped up in the meantime And again some more in the last 20 minutes #NotDeadYet |
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
You can ask Veneto how to do that modification to block beta work. But I only got one beta and since then nothing but clean 4.20. So is it worth the effort to mess around? Sorry about that..couldn't remember your nickname and was in a rush this morning to get out of the house. I thought he was asking how to block the Beta app so he wouldn't get that work, being that it errors out right now. Is that even possible? |
Greg_BE Send message Joined: 30 May 06 Posts: 5690 Credit: 5,859,226 RAC: 14 |
Wow, the queue got eaten up that fast? There must be some people that take 100 tasks at a time. |
Message boards :
Number crunching :
Problems and Technical Issues with Rosetta@home
©2024 University of Washington
https://www.bakerlab.org