Hi,

in the last few weeks I always get computation errors like this:

2007-09-30 14:04:38 [SETI@home] Restarting task 07mr07ak.6690.11933.7.6.237_1 using setiathome_enhanced version 527
2007-09-30 14:06:03 [SETI@home] Deferring communication for 1 min 0 sec
2007-09-30 14:06:03 [SETI@home] Reason: Unrecoverable error for result 07mr07ak.6690.11933.7.6.237_1 (process exited with code 131 (0x83, -125))

I also had these problems with the "original" version 5.23 of S@H client.

The second problem I have is that the computer crashes from time to time if I run S@H. When I stop S@H in BOINC the computer runs without problems.

My system is a PPC 1.8 GHz G4 Mac, OS X 10.4.10, 896MB RAM

Does anybody know what is going wrong here?

Kind regards,
Andreas

---

It looks like a hardware problem.
Do you Mac hang complete if the crash happening ?
Could you please let a hardware diagnostic programm on your Mac run to verify that the CPU, Memory and cache is OK.

---

It looks like a hardware problem.
Do you Mac hang complete if the crash happening ?
Could you please let a hardware diagnostic programm on your Mac run to verify that the CPU, Memory and cache is OK.

Yes, my computer hangs completely when the crash happens.
I've run the hardware test rogram of Apple several times on simple level without any problem. On extended level I got error report when testing the mainboard already. This happens may be due to installed CPU acceleration board. So I cannot clearly say everything is ok. The only thing I changed on hardware level of my Mac was that I exchanged the RAM modules of socket 1 and 3 with each other. I have bought my G4 with standard 128MB module (socket 1) plus an additional 256MB module (socket 2). I extended the RAM then later with a 512MB module (socket 3). A few months ago I exchanged the 128MB and the 512MB module with each other.

On the other hand I'm running Einstein@Home without any troubles or problems.

---

I get similar errors with every seti unit since I started using Boinc with multiple projects. The other projects don't fail.

Currently Boinc 5.10.20 Seti Enhanced 5.27

Other Projects Einstein and Climate Prediction. All projects 100 each. Multi core CPU All projects set top Processor.

---

I've run the hardware test rogram of Apple several times on simple level without any problem. On extended level I got error report when testing the mainboard already. This happens may be due to installed CPU acceleration board. So I cannot clearly say everything is ok. The only thing I changed on hardware level of my Mac was that I exchanged the RAM modules of socket 1 and 3 with each other. I have bought my G4 with standard 128MB module (socket 1) plus an additional 256MB module (socket 2). I extended the RAM then later with a 512MB module (socket 3). A few months ago I exchanged the 128MB and the 512MB module with each other.

Have you asked at apple or a apple dealer about this problem ?
I think this behavior is not normal, and this crashs and behavior looks like at a hardware problem.

Btw. have you tried to diag this problem with other freeware tools ?

On the other hand I'm running Einstein@Home without any troubles or problems.

I think, that the einstein application complete utilize the hardware, so it could be that this applications ran without any problems.
I had such a behavior at SIMAP and SETI. All projects works successfull, but the SIMAP and SETI WUs crashed. A bad memory module was the cause...

---

I get similar errors with every seti unit since I started using Boinc with multiple projects. The other projects don't fail.

Currently Boinc 5.10.20 Seti Enhanced 5.27

Other Projects Einstein and Climate Prediction. All projects 100 each. Multi core CPU All projects set top Processor.

What errors excatly do you get ?
Can you please post the errror messages and some stderr.txt from some crashed results here ? - I can't look at your computers, because they are hidden.

---

Have you asked at apple or a apple dealer about this problem ?
I think this behavior is not normal, and this crashs and behavior looks like at a hardware problem.

Btw. have you tried to diag this problem with other freeware tools ?

So far I didn't ask apple for help. If you can tell me the names of the freeware tools you mentioned, I will try to diag this problem.

I'm not sure if the Apple test tool does work correctly with my accelerated Mac. It says at one point that all installed extras have to be removed. I think I would have to replace the accelerator board with the original CPU, this would be to much work for me.
The behaviour is also strange to me because a few months ago I didn't had this massive problems with S@H.

On the other hand I'm running Einstein@Home without any troubles or problems.

I think, that the einstein application complete utilize the hardware, so it could be that this applications ran without any problems.
I had such a behavior at SIMAP and SETI. All projects works successfull, but the SIMAP and SETI WUs crashed. A bad memory module was the cause...

I will put the RAM modules back to the first configuration I had and see if the problems still are there.

---

I get similar errors with every seti unit since I started using Boinc with multiple projects. The other projects don't fail.

Currently Boinc 5.10.20 Seti Enhanced 5.27

Other Projects Einstein and Climate Prediction. All projects 100 each. Multi core CPU All projects set top Processor.

What errors excatly do you get ?
Can you please post the errror messages and some stderr.txt from some crashed results here ? - I can't look at your computers, because they are hidden.

This is part of the output of stdoutdae.txt:
2007-09-30 09:25:31 [SETI@home] [file_xfer] Started download of file 07mr07ak.6690.11933.7.6.237
2007-09-30 09:25:52 [SETI@home] [file_xfer] Finished download of file 07mr07ak.6690.11933.7.6.237
2007-09-30 09:25:52 [SETI@home] [file_xfer] Throughput 18080 bytes/sec
2007-09-30 10:22:50 [SETI@home] Starting 07mr07ak.6690.11933.7.6.237_1
2007-09-30 10:22:50 [SETI@home] Starting task 07mr07ak.6690.11933.7.6.237_1 using setiathome_enhanced version 527
here the crash happens, afterwards I restarted my computer
2007-09-30 11:15:09 [---] Starting BOINC client version 5.10.20 for powerpc-apple-darwin
then a little later
2007-09-30 14:04:38 [SETI@home] Restarting task 07mr07ak.6690.11933.7.6.237_1 using setiathome_enhanced version 527
2007-09-30 14:06:03 [SETI@home] Deferring communication for 1 min 0 sec
2007-09-30 14:06:03 [SETI@home] Reason: Unrecoverable error for result 07mr07ak.6690.11933.7.6.237_1 (process exited with code 131 (0x83, -125))
2007-09-30 14:06:03 [SETI@home] Computation for task 07mr07ak.6690.11933.7.6.237_1 finished
again somewhat later
2007-09-30 14:28:03 [SETI@home] [file_xfer] Started download of file 07mr07ak.25902.9070.11.6.55
2007-09-30 14:28:09 [SETI@home] [file_xfer] Finished download of file 07mr07ak.25902.9070.11.6.55
2007-09-30 14:28:09 [SETI@home] [file_xfer] Throughput 69763 bytes/sec
2007-09-30 15:06:26 [SETI@home] Starting 07mr07ak.25902.9070.11.6.55_0
2007-09-30 15:06:26 [SETI@home] Starting task 07mr07ak.25902.9070.11.6.55_0 using setiathome_enhanced version 527
here again the computer crashed
2007-09-30 15:10:24 [---] Starting BOINC client version 5.10.20 for powerpc-apple-darwin

And here is the output of the crash reporter log
Host Name: Andis-Mac
Date/Time: 2007-09-30 14:06:01.435 +0200
OS Version: 10.4.10 (Build 8R218)
Report Version: 4

Command: seti_enhanced-ppc-v7.1mb-g4-nographics
Path: ../../projects/setiathome.berkeley.edu/seti_enhanced-ppc-v7.1mb-g4-nographics
Parent: boinc [201]

Version: ??? (???)

PID: 649
Thread: 0

Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x00000000

Thread 0 Crashed:
0 ...ed-ppc-v7.1mb-g4-nographics 0x00165370 t2bv_16 + 1280
1 <<00000000>> 0x00000010 0 + 16
2 ...ed-ppc-v7.1mb-g4-nographics 0x001dad5c apply + 140
3 ...ed-ppc-v7.1mb-g4-nographics 0x0000555c seti_analyze(ANALYSIS_STATE&) + 3212
4 ...ed-ppc-v7.1mb-g4-nographics 0x0001ea68 seti_do_work() + 48
5 ...ed-ppc-v7.1mb-g4-nographics 0x0001fe90 worker() + 136
6 ...ed-ppc-v7.1mb-g4-nographics 0x0000be4c main + 624
7 ...ed-ppc-v7.1mb-g4-nographics 0x0000294c _start + 760
8 ...ed-ppc-v7.1mb-g4-nographics 0x00002650 start + 48

Thread 1:
0 libSystem.B.dylib 0x900411f8 mach_wait_until + 8
1 libSystem.B.dylib 0x90040fc4 nanosleep + 388
2 libSystem.B.dylib 0x90040df0 sleep + 144
3 ...ed-ppc-v7.1mb-g4-nographics 0x00092b60 boinc_sleep(double) + 48
4 ...ed-ppc-v7.1mb-g4-nographics 0x0008d638 timer_thread(void*) + 40
5 libSystem.B.dylib 0x9002bd08 _pthread_body + 96

Thread 0 crashed with PPC Thread State 64:
srr0: 0x0000000000165370 srr1: 0x000000000200f030 vrsave: 0x00000000ffffffff
cr: 0x44424482 xer: 0x0000000020000004 lr: 0x0000000000000010 ctr: 0x000000000000007e
r0: 0x000000000403e020 r1: 0x00000000bffff580 r2: 0x000000000403f020 r3: 0x00000000bffff6a0
r4: 0x0000000004039020 r5: 0x00000000004ae520 r6: 0x00000000bffff690 r7: 0x000000000403c820
r8: 0x000000000403f820 r9: 0x000000000403d020 r10: 0x00000000004ae460 r11: 0x0000000000000000
r12: 0x0000000004038020 r13: 0x0000000000000010 r14: 0x000000000403d820 r15: 0x000000000403e820
r16: 0x000000000403b020 r17: 0x000000000403a820 r18: 0x0000000004039820 r19: 0x000000000403b820
r20: 0x0000000004038820 r21: 0x000000000403a020 r22: 0x000000000403c020 r23: 0x00000000bffff6a0
r24: 0x00000000bffff690 r25: 0x00000000bffff6a0 r26: 0x0000000000000000 r27: 0x00000000bffff5c0
r28: 0x00000000bffff5a0 r29: 0x00000000bffff5b0 r30: 0x0000000000000200 r31: 0x00000000004ae5a0

If you need more information, please let me know

---

Do you have better success, after put the RAM modules back ? Is it running stable now ?

---

Do you have better success, after put the RAM modules back ? Is it running stable now ?

No, I still have the same problems.

---

Have you asked your Apple dealer or tht Apple support about the problems reported from test tools about the acclerator board ?
What happens if you use the actual offical SETI application (version 5.28) ?

---

Have you asked your Apple dealer or tht Apple support about the problems reported from test tools about the acclerator board ?
What happens if you use the actual offical SETI application (version 5.28) ?

I haven't asked Apple support, because the Apple test tool expects the original hardware. At least some information in the tool says that. So I'm not sure if Apple support could help me. Well to be sure if the problem is related to the hardware I would have to remove the accelerator CPU board and replace it with the original CPU. I'd like not to do this to prevent any damage to it.
I had the same problems using the official SETI application, so far I'm not sure if I already used 5.28.

You mentioned some freeware test tools some time ago. Could you tell me the names of the tools and where to find?

---

Have you asked your Apple dealer or tht Apple support about the problems reported from test tools about the acclerator board ?
What happens if you use the actual offical SETI application (version 5.28) ?

I haven't asked Apple support, because the Apple test tool expects the original hardware. At least some information in the tool says that. So I'm not sure if Apple support could help me. Well to be sure if the problem is related to the hardware I would have to remove the accelerator CPU board and replace it with the original CPU. I'd like not to do this to prevent any damage to it.
I had the same problems using the official SETI application, so far I'm not sure if I already used 5.28.

I would recommend you to try the offical 5.28, too. I think, there is an issue with your system, but I did not know where it is in the moment.

You mentioned some freeware test tools some time ago. Could you tell me the names of the tools and where to find?

Rember is a good memory test tool : http://www.apple.com/downloads/macosx/system_disk_utilities/rember.html
And prime95 do a stress test of your system.

---

I would recommend you to try the offical 5.28, too. I think, there is an issue with your system, but I did not know where it is in the moment.

I updated to latest BOINC (5.10.29) and also S@H client (5.28), still the problem is there. The computer crashes from time to time and usually the work units are not calculated correctly by my Mac. Just look at my computers results. Either S@H client has crashed or I get this message: SETI@Home Informational message -9 result_overflow NOTE: The number of results detected exceeds the storage space allocated.
Do you know which storage space is ment? Is it the RAM or is it the harddrive?
For now I have stopped S@H calculation because of the system crashes.

Rember is a good memory test tool : http://www.apple.com/downloads/macosx/system_disk_utilities/rember.html
And prime95 do a stress test of your system.

Thanks for the info. If loaded rember and run some tests without any negative results so far. When I have time I will run it in SingelUser mode.
Thanks for your replys so far.

---

I updated to latest BOINC (5.10.29) and also S@H client (5.28), still the problem is there. The computer crashes from time to time and usually the work units are not calculated correctly by my Mac. Just look at my computers results. Either S@H client has crashed or I get this message: SETI@Home Informational message -9 result_overflow NOTE: The number of results detected exceeds the storage space allocated. Do you know which storage space is ment? Is it the RAM or is it the harddrive?

The “–9 overflow” messages do not normally indicate a problem: they usually come from WUs that have excessive noise or strong terrestrial signals in them. (There’s a military radar installation in Puerto Rico that’s been identified as a frequent contributor of the latter.) Here “storage space” just refers to the maximum size of a result file to be uploaded, nothing to do with your RAM or HD.

That said, since your quorum partners aren’t getting the same results—your tasks with the –9 message appear to be coming up invalid—the messages are most likely symptomatic of a problem in your case, that even when it gets through a task without obvious computation errors your system is finding spurious signals, and could well be related to the crashes.

Have you tried reverting to an older BOINC version? 5.4.9 is working well for me. If your other projects are doing fine, it doesn’t seem very likely that BOINC is at fault, but you never know. The problems I’ve had on my G4/733 seem to be caused by the screensaver graphics, so Einstein@home is definitely one of the projects affected—and S@h isn’t, because I run Alex Kan’s optimized app without graphics here.

---

The error -9 could also indicate a hardware (CPU, Memory or Cache) problem, or some problems with the SETI application. But, as all your results get no validated and produce -9's at every result, I think that it is hardware related problem.
I recommend you to download prime95 and make a stress test on the system.

---