Still having issues with 5.25. Same problems as I posted before, and apparently, many others are posting on. All problems with previous versions were minor and few in number. I'm running it on 6 hosts, all with varying CPU/OS's. All with plenty of RAM. Anxiously awaiting some fixes =)
Welcome all newcommers.

---

I have a graphical problem with all the workunits. I have a laptop with widescreen (1280*800), and when I see the graphics in full screen or using the screensaver it is cropped and the bottom line is missing. Here is a screenshoot:


It only happens when I suspend a WU and resume it. When a new WU starts it fits the resulotion, but if I exit the screensaver and let it start again, the graphics don't fit anymore.

This happens on Mac OS X too. It's related to widescreen displays. It may be related to using the default boinc graphics code which was written for normal aspect ratio displays. See http://www.ssl.berkeley.edu/pipermail/boinc_dev/2006-July/006034.html

and the following messages in the thread.

Tim

---

Rosetta 5.25 on Mac OS X 10.4.7 incorrectly classifies the target t314__CASP7_FOLLOWUP_ABRELAX_SAVE_ALL_OUT_BARCODE_perfectss__1066_42278_0
as unknown, even though the display clearly shows it as known.

Use the link to email me.

---

Well i've waited as long as I care to wait spending my $ contributing.

PS.... 5.25 has been the worst release, it has made me quit when I spent most of my time trying to recuit people to Rosetta. I posted months ago on the issue... nothing.

Now I retire Rosetta from my machines, and can enjoy my lower power bill too. 5 machines were eating plenty.

---

Well i've waited as long as I care to wait spending my $ contributing.

PS.... 5.25 has been the worst release, it has made me quit when I spent most of my time trying to recuit people to Rosetta. I posted months ago on the issue... nothing.

Now I retire Rosetta from my machines, and can enjoy my lower power bill too. 5 machines were eating plenty.

That is sad. For most people it seems 5.25 works quite well including me. I had this "sit with 0% CPU usage" only over at Ralph with some pre 5.25 version. Have you ever attached to RALPH to see whether you have the same problems there as here?

I think we'll see new versions soon. Now that CASP is over they probably resume tweaking the app.

---

I have had ZERO problems with this release of Rosetta. I have Rosetta installed on 7 computers and it runs on every computer just fine with no problems. I am running the BOINC version 5.4.9.

---

Perhaps all of you who think 5.25 is flawless need to read this thread. Nothing but posts of issues. All the same for every person. I posted a month+ ago about it, nothing was said in any way. I posted again a few weeks ago, still nothing. Now I quit and people notice. I've added somewhere about 40 machines to this project because I think it's a great cause, a great idea, and a new path.
So I lost who knows how many hours with 5.25 stopping at 99-100%, then dumping the WU on reboot. It's just frustrating. If I didn't have so many boxes to check all the time to make sure they were running, It wouldn't bother me. I realize the attention has been focused on CASP, but there is MORE than enough people bringing forth issues since 5.25's debut. And for the record, I had no problems UNTIL 5.25.

---

I have had ZERO problems with this release of Rosetta. I have Rosetta installed on 7 computers and it runs on every computer just fine with no problems. I am running the BOINC version 5.4.9.

I have to restart 1 or 2 machines each week. But that's not bad for 22. I check my stats every day to make sure all of them are connecting. When one of them gets late, it means I have to restart the software.

If I didn't have to do that my efficiency would be higher, but not significantly. What can I provide to help figure out what is happening?

---

Team Starfire World BOINC

I don't know what happened to resultid=32044316 but it failed, sped through 3 waiting work units at the speed of light, and thoroughly mixed up the computer.

---

Sangamon I would think it would be helpful if you could note any patterns in your farm. Is it the same 2 or 3 boxes getting hung up? Or random? Do they all run same BOINC version? Same OS udpates? Same BOINC Preferences and locations?

You might make it easy on yourself, whenever a machine gets hung up, attach it to Ralph for like a 10% resource share. So you can just see if the machine is already on Ralph, and start to get a feel if there are perfect machines in your farm, or if all are effected.

Then screutenize the Ralph work done closely. If failures occur on Ralph, more diagnostic data is returned to the project. Post on the boards there the specific WUs you see hanging up or crashing BOINC.

If you have specific Rosetta WUs that caused problems, post their IDs here. Sometimes there are issues with how the WUs are created. Other times it's a specific random number for a model that uncovers a problem.

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

The people who have had no problems, I believe have been lucky and are probably running Windows. I have 4 computers at home, 2 do Windows and 2 do Linux. One of the Windows machines has had about 2 or 3 Wu's lock up but that is all. The 2 Linux machines have had numerous WU's lockup (around 2 dozen (24) at least). In all cases the "CPU time" and "To completion" times stop and the host CPU drops back to idle but does not move from that WU, saying it is running when in fact it is not, even after many hours to days.
Suspending the WU and resuming has no effect, restarting the Boinc Client/Manager has no effect, only a reboot will get the unit working again. More often than not after restarting the WU will error out anyway.
I do not intend to reboot every time a WU stops so at this stage I have been aborting the WU's in question to keep my computer doing something useful.
Seems to be more a Linux problem than Windows and mostly I get a "Segmentation Error" as the cause of the failure. The time to failure can be from a few minutes/percent to almost complete (80 to 90% done).
Happens on both Rosetta@home and also on Ralph@home, with more failures on Ralph than on Rosetta.

---

The people who have had no problems, I believe have been lucky and are probably running Windows. I have 4 computers at home, 2 do Windows and 2 do Linux. One of the Windows machines has had about 2 or 3 Wu's lock up but that is all. The 2 Linux machines have had numerous WU's lockup (around 2 dozen (24) at least).

My linux crunchers have been going 24/7 with no problems for a long time now. (I've had 17 Linux crunchers going 100% on Rosetta during casp, but have now dropped back to 13 Linux crunchers.)

I'm using the standard BOINC 5.2.13.

I've detached from rosetta for the time, and moved to 100% share on ralph. I don't want to just give up, 5.25 is just making me loopy.

---

Still getting same problems on my 2 Linux machines, Windows machines ok.
WU says it is running but nothing is happening, no timers are moving and CPU is idle. Will not switch to another project as per preferences, my only solution is to abort as I can not be rebooting all the time to restart WU.
Getting "process exited with code 131" ERROR:SIGSEGV:segmentation violation"
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28130067
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28130003 (aborted)
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129983
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129982
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129975 (aborted)
Plus these 2 the day before (12th)
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=27385306 (aborted)
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=27385307 (aborted)
Also have been getting the same error/problems on Ralph, again with Linux.

---

process exited with code 131 (0x83) here
Boinc 5.4.9 on Linux Suse 10.0.

---

Arnaud

Hello Conan...would appreciate knowing where you get that download for the core client 5.5.0 that you are using....you are getting the same credits in ROSETTA using that client as I did using 5.3.12, but not 5.4.11....thanks, Stan

user stats

---

Still getting same problems on my 2 Linux machines, Windows machines ok.
WU says it is running but nothing is happening, no timers are moving and CPU is idle. Will not switch to another project as per preferences, my only solution is to abort as I can not be rebooting all the time to restart WU.
Getting "process exited with code 131" ERROR:SIGSEGV:segmentation violation"
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28130067
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28130003 (aborted)
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129983
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129982
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129975 (aborted)
Plus these 2 the day before (12th)
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=27385306 (aborted)
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=27385307 (aborted)
Also have been getting the same error/problems on Ralph, again with Linux.

A follow up to the above, all of the new Wu's that I have processed today have died. All with the same errors as above and here are another 2
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28130084
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=28129985

2 have died at 3741 and 3758 seconds (wu 28130067,28129983), with the others going between 5347 and 5400 seconds (about 1 1/2 hours). Coincidently my switch time between projects is 90 minutes (1 1/2 hours).
The WU's either error out at the times mentioned or hang with nothing happening. This problem has been increasing over the last few days.
The 2 Linux machines having most of the problems are both AMD, both dual CPUs, one a 275 with 4 GB RAM and one a 848 with 2 GB RAM, both running Linux Fedora Core 3.

---

Rosetta 5.25 writes its checkpoints before calculating, if it shouldn't stop now. With the actual betas of boinc this leads to the funny situation, that boinc waits for a checkpoint to switch task - sees the checkpoint - switches task - and Rosetta sits there with 100% done, till the other projects got their share, before the result is uploaded.

Norbert

---

Another stuck work unit:

2f21X_BOINC_ABRELAX_SAVE_ALL_OUT_BARCODE__1075_31308
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=27912763

It has been running for more than 2 days but accumulated only 6 hours of CPU time and is stuck at 74.4%.

When I stopped BOINC, the Rosetta processes did not terminate.
I rebooted the machine and the WU immediately terminated with the error

ERROR:: Exit at: initialize.cc line:1618

Hi,

I'm having the same problems with Rosetta@Home "hanging" (it shows "running" but the CPU is at 0%). Usually it occurs within 23%~26% of processing the unit. The same thing happens with World Community Grid as well, but I know that's another project. ALL other projects work fine. I'm on a P4C 2.6 GHz, 512 MB RAM running Xubuntu. Nothing is overclocked. The workunits below I know are "stuck":

FRA_t370_CASPR_hom001_6_t370_4_2a2jA_IGNORE_THE_REST_223_1078_61_0
FRA_t322_CASPR_hom001_6_t322_3_1u1zA_IGNORE_THE_REST_17_1079_65_1

There are a lot more that I've had to abort over the weeks, but my log only goes so far.

---