Message boards : Number crunching : Minirosetta 3.59
Author | Message |
---|---|
David E K Volunteer moderator Project administrator Project developer Project scientist Send message Joined: 1 Jul 05 Posts: 1018 Credit: 4,334,829 RAC: 0 |
This version includes a relatively recent sync with the Rosetta source. New and updated protocols and code refactoring efforts have been made in this version since it has been quite a while since the last update. These include protocols for modeling cyclic peptides, ab initio modeling starting from an extended chain in the hybridize protocol (RosettaCM), etc. Please report issues regarding this update in this thread. |
TPCBF Send message Joined: 29 Nov 10 Posts: 111 Credit: 4,991,034 RAC: 879 |
Those WUs seem to have some serious problems, at least here on my Windows 8.1 laptop (4core i3, 8GB of RAM) They just rack up a few minutes of CPU time before kind of stopping, just "real" time advancing but with no apparent processing happening for hours... And when restarting BOINC, they might pretend to work for a (very) short time, then crapping out with "Computation Error"... Ralf |
P . P . L . Send message Joined: 20 Aug 06 Posts: 581 Credit: 4,865,274 RAC: 0 |
Yea, the graphics window finally works under Ubuntu after how many years I've lost count. As for the tasks, I haven't seen any problems with the ones running on my rigs. |
David E K Volunteer moderator Project administrator Project developer Project scientist Send message Joined: 1 Jul 05 Posts: 1018 Credit: 4,334,829 RAC: 0 |
Those WUs seem to have some serious problems, at least here on my Windows 8.1 laptop (4core i3, 8GB of RAM) Are you seeing this problem with all WUs? There doesn't seem to be a general issue as far as I can tell so far. |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
Those WUs seem to have some serious problems, at least here on my Windows 8.1 laptop (4core i3, 8GB of RAM) Their first WU seems to have failed with an error about getting lock on lockfile: https://boinc.bakerlab.org/rosetta/result.php?resultid=739734418 Starting work on structure: _00001 [2015- 6-15 16: 8:27:] :: BOINC:: Initializing ... ok. Can't acquire lockfile - exiting FILE_LOCK::unlock(): close failed.: No error [2015- 6-15 16: 9:10:] :: BOINC:: Initializing ... ok. Can't acquire lockfile - exiting FILE_LOCK::unlock(): close failed.: No error ...and it repeats about every 45 seconds. Rosetta Moderator: Mod.Sense |
TPCBF Send message Joined: 29 Nov 10 Posts: 111 Credit: 4,991,034 RAC: 879 |
Not all, but still weird. 6 other tasks seemed to have finished today fine, though I see messages about "exit status 0" and "you might have to reset the project if this continues" in the logs.Those WUs seem to have some serious problems, at least here on my Windows 8.1 laptop (4core i3, 8GB of RAM) Ralf |
nanoprobe Send message Joined: 5 Apr 09 Posts: 8 Credit: 381,804 RAC: 0 |
Just started running this project again after a long hiatus. The tasks I'm getting are taking about 24 hours each on a 2600k. Is that normal? |
SDotloe Send message Joined: 16 Mar 15 Posts: 1 Credit: 45,350,100 RAC: 0 |
Just started running this project again after a long hiatus. The tasks I'm getting are taking about 24 hours each on a 2600k. Is that normal? Might be your cpu runtime preference. Otherwise, they shouldn't run that long. |
nanoprobe Send message Joined: 5 Apr 09 Posts: 8 Credit: 381,804 RAC: 0 |
Just started running this project again after a long hiatus. The tasks I'm getting are taking about 24 hours each on a 2600k. Is that normal? Runtime preferences are 100% 24/7/365 |
David E K Volunteer moderator Project administrator Project developer Project scientist Send message Joined: 1 Jul 05 Posts: 1018 Credit: 4,334,829 RAC: 0 |
Just started running this project again after a long hiatus. The tasks I'm getting are taking about 24 hours each on a 2600k. Is that normal? That was me accidentally logged in as SDotloe. I meant the CPU runtime preference. When you login, go to your preferences, particularly the "Resource share and graphics: Rosetta@home preferences". There's a "Target CPU run time" option. I wonder if that is set to 24 hours. If so, set it to your preferable run time. If it is not set and is the default then maybe you have some long running work units. |
nanoprobe Send message Joined: 5 Apr 09 Posts: 8 Credit: 381,804 RAC: 0 |
Just started running this project again after a long hiatus. The tasks I'm getting are taking about 24 hours each on a 2600k. Is that normal? CPU target runtime is set to 1 day and the tasks are taking 1 day. Is there any correlation there? I'm going to change it and see what happens. |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
CPU target runtime is set to 1 day and the tasks are taking 1 day. Is there any correlation there? I'm going to change it and see what happens. Yep. The only other "catch" is that you can set it differently for each "venue" if you like. So you have to make the change to the settings for the venue that the PC is configured to be a part of. Rosetta Moderator: Mod.Sense |
nanoprobe Send message Joined: 5 Apr 09 Posts: 8 Credit: 381,804 RAC: 0 |
CPU target runtime is set to 1 day and the tasks are taking 1 day. Is there any correlation there? I'm going to change it and see what happens. Changed CPU runtime to 4 hours and tasks are now taking 4-5 hours. Thanks for the help. Curious as to why you have a CPU runtime option? I don't think I've seen that on any other project that I've run. |
Timo Send message Joined: 9 Jan 12 Posts: 185 Credit: 45,649,459 RAC: 0 |
Curious as to why you have a CPU runtime option? I don't think I've seen that on any other project that I've run. Great question. Setting a longer run-time actually means that each WU you crunch will run more decoys (more searches through different paths of confirmation-space) for the same core model/protein/data set. Many thousands of decoys are needed for obtaining an accurate prediction of a single protein. As an example only; Setting a short runtime like 3 hours may compute 2-3 decoys for a given target model. Setting a longer runtime like 10 hours would compute say 8-12 decoys for the same target model. Either way, thousands of computations are needed for a given target, so what is the advantage? Well, there are hundreds of different core models running on Rosetta at any given time, so setting a longer runtime means that your WU will work with the SAME core model for a bit longer, and this means that some core datasets can be reused by the same WU. This all results in saving some bandwidth / less pestering of the project server/less overhead switching out the core targets and datasets, etc. It's akin to telling the computer to concentrate on a certain task for longer rather than jumping between tasks and having to load new datasets/bug the server/etc. Thus, longer runtimes mean less overhead and more efficiency. |
sgaboinc Send message Joined: 2 Apr 14 Posts: 282 Credit: 208,966 RAC: 0 |
i'm doing a first run via 3.59 on opensuse linux 13.2, thanks much for adding the show graphics feature in 3.59 for linux, this is truly a cool/great feature :) |
TPCBF Send message Joined: 29 Nov 10 Posts: 111 Credit: 4,991,034 RAC: 879 |
Ok, here we go again. Yesterday, two WU crapped out after stopping and restarting BOINC a couple of times with computation error. Today (actually, since last night) there are again two WUs that start fine, then just rack up "real time" without doing anything past the last check point in terms of CPU time, just blocking two cores on this CPU for no good reason at all...Not all, but still weird. 6 other tasks seemed to have finished today fine, though I see messages about "exit status 0" and "you might have to reset the project if this continues" in the logs.Those WUs seem to have some serious problems, at least here on my Windows 8.1 laptop (4core i3, 8GB of RAM) Ralf |
Ryan Send message Joined: 17 Aug 09 Posts: 2 Credit: 5,143,774 RAC: 0 |
Ever since the new update launch, BOINC freezes during the transition from one Rosetta Mini app to another. The only way I have found to stop this is to suspend one app to allow the other to run. This is becoming rather frustrating, because instead of running in the background I have to continually monitor the new downloads to the system. Even if it is not or cannot fixed I am going to continue to run this appilication as I believe it. Thanks, Ryan |
Ryan Send message Joined: 17 Aug 09 Posts: 2 Credit: 5,143,774 RAC: 0 |
Ok, here we go again. Yesterday, two WU crapped out after stopping and restarting BOINC a couple of times with computation error. Today (actually, since last night) there are again two WUs that start fine, then just rack up "real time" without doing anything past the last check point in terms of CPU time, just blocking two cores on this CPU for no good reason at all...Not all, but still weird. 6 other tasks seemed to have finished today fine, though I see messages about "exit status 0" and "you might have to reset the project if this continues" in the logs.Those WUs seem to have some serious problems, at least here on my Windows 8.1 laptop (4core i3, 8GB of RAM) This is exactly what is happening to me!!! |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
Please report which BOINC version is installed and which host you are making yout observations on. Rosetta Moderator: Mod.Sense |
TPCBF Send message Joined: 29 Nov 10 Posts: 111 Credit: 4,991,034 RAC: 879 |
Please report which BOINC version is installed and which host you are making yout observations on.BOINC 7.4.42 on Windows 8.1/64(4core i3/8GB of RAM). Task ID 742059437 and Task ID 742059580. After restarting BOINC 3 or 4 times and shutdown and restart my laptop half a couple times since yesterday, right now both WUs are currently continuing working again, but I have the feeling that they will stop and do the same thing all over again in a while... Ralf |
Message boards :
Number crunching :
Minirosetta 3.59
©2024 University of Washington
https://www.bakerlab.org