I'm not remembering the whole thread (and it's way too early/late - depending on how you look at it - to go back through it).

Is there something that seems to trigger the crash? If there is, you could always run it under gdb. I've only done it once, and with help, so I'm not any use beyond the suggestion.

...

It seems, but only seems, that it crashes after a hydrogen@home work unit has been downloaded. But I cannot believe downloading a work unit would do it. It seems, but only seems, that the hydrogen@home work unit does not get started. But this could be because it starts and causes the crash before the boinc client gets around to logging that it started. Many hydrogen@home work units complete, many of them successfully.

Well I have now run both machines for a little over a week with hydrogen@home getting no work units and the clients have not crashed. I will try for another week, but it is suggestive...

---

It's now happening to others as well.

---

I'm not remembering the whole thread (and it's way too early/late - depending on how you look at it - to go back through it).

Is there something that seems to trigger the crash? If there is, you could always run it under gdb. I've only done it once, and with help, so I'm not any use beyond the suggestion.

...

It seems, but only seems, that it crashes after a hydrogen@home work unit has been downloaded. But I cannot believe downloading a work unit would do it. It seems, but only seems, that the hydrogen@home work unit does not get started. But this could be because it starts and causes the crash before the boinc client gets around to logging that it started. Many hydrogen@home work units complete, many of them successfully.

Well I have now run both machines for a little over a week with hydrogen@home getting no work units and the clients have not crashed. I will try for another week, but it is suggestive...

It has now gone for about three weeks without downloading any work for hydrogen@home and has not crashed on either of my machines.

With the arrival of client 6.2.15, I turned on hydrogen@home for about a day and it seems to be OK. I do not draw any conclusion yet, but no crashes.

I do not know if hydrogen@home has been fixed, or if the latest boinc client has been made more resistant to whatever hydrogen@home has been doing.

---

I'm not remembering the whole thread (and it's way too early/late - depending on how you look at it - to go back through it).

Is there something that seems to trigger the crash? If there is, you could always run it under gdb. I've only done it once, and with help, so I'm not any use beyond the suggestion.

I have been running the boinc client 6.2.15 for a bit over a week now, allowing hydrogen@home tasks, and it just crashed. Here is the end of my trillian.boinc.log file:

24-Aug-2008 06:46:16 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 26235 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:46:21 [Hydrogen@Home] Scheduler request succeeded: got 1 new tasks
24-Aug-2008 06:46:23 [Hydrogen@Home] Started download of 1219574772_nsc12157.mol2
24-Aug-2008 06:46:23 [Hydrogen@Home] Started download of 1219574772_pdb1uwu.pdb
24-Aug-2008 06:46:25 [Hydrogen@Home] Finished download of 1219574772_nsc12157.mol2
24-Aug-2008 06:46:25 [Hydrogen@Home] Started download of ad_nsc12157.mol2_pdb1uwu.pdb_1219574772
24-Aug-2008 06:46:26 [Hydrogen@Home] Finished download of 1219574772_pdb1uwu.pdb
24-Aug-2008 06:46:26 [Hydrogen@Home] Finished download of ad_nsc12157.mol2_pdb1uwu.pdb_1219574772
24-Aug-2008 06:46:31 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 27369 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:46:36 [Hydrogen@Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:46:36 [Hydrogen@Home] Message from server: No work sent
24-Aug-2008 06:47:06 [Predictor @ Home] Sending scheduler request: To fetch work. Requesting 26238 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:47:11 [Predictor @ Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:47:36 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 27369 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:47:41 [Hydrogen@Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:47:41 [Hydrogen@Home] Message from server: No work sent
24-Aug-2008 06:48:41 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 26243 seconds of work, reporting 1 completed tasks
24-Aug-2008 06:48:46 [Hydrogen@Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:48:46 [Hydrogen@Home] Message from server: No work sent
24-Aug-2008 06:49:46 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 26245 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:49:51 [Hydrogen@Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:49:51 [Hydrogen@Home] Message from server: No work sent
24-Aug-2008 06:50:52 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 26241 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:50:57 [Hydrogen@Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:50:57 [Hydrogen@Home] Message from server: No work sent
24-Aug-2008 06:52:52 [Hydrogen@Home] Sending scheduler request: To fetch work. Requesting 26241 seconds of work, reporting 0 completed tasks
24-Aug-2008 06:52:57 [Hydrogen@Home] Scheduler request succeeded: got 0 new tasks
24-Aug-2008 06:52:57 [Hydrogen@Home] Message from server: No work sent

Then the boinc client exited, killing all its children.

Here is the end of the trillian.error.log.

SIGSEGV: segmentation violation
Stack trace (8 frames):
/home/boinc/BOINC/boinc[0x8094e76]
[0x3b4420]
/home/boinc/BOINC/boinc[0x8066c52]
/home/boinc/BOINC/boinc[0x805d6d4]
/home/boinc/BOINC/boinc[0x80840e2]
/home/boinc/BOINC/boinc[0x80842cb]
/lib/libc.so.6(__libc_start_main+0xdc)[0x49cdec]
/home/boinc/BOINC/boinc(__gxx_personality_v0+0x195)[0x804bda1]

Exiting...

The machine is not too hot:

Sun Aug 24 07:00:01 EDT 2008
w83627hf-isa-0290
Adapter: ISA adapter
VCore: +1.47 V (min = +1.36 V, max = +1.47 V)
+3.3V: +3.31 V (min = +3.14 V, max = +3.46 V)
VBat: +3.18 V (min = +2.40 V, max = +3.60 V)
+5V: +4.97 V (min = +4.84 V, max = +5.24 V)
+12V: +12.04 V (min = +11.49 V, max = +12.59 V)
-12V: -11.70 V (min = -13.02 V, max = -11.37 V)
V5SB: +5.43 V (min = +4.84 V, max = +5.24 V)
CPU0 fan: 2410 RPM (min = 1592 RPM, div = 8)
CPU1 fan: 2058 RPM (min = 1592 RPM, div = 8)
System: +38 C (high = +54 C, hyst = +52 C) sensor = thermistor
CPU0: +38.5 C (high = +60 C, hyst = +58 C) sensor = thermistor
CPU1: +39.0 C (high = +60 C, hyst = +58 C) sensor = thermistor
vid: +1.525 V (VRM Version 9.0)

I also run climate prediction, setiathome, rosetta, world community grid, malaria control, and predictor (that does not usually get any work), with no problems.

I still imagine there is a problem with hydrogen at home, but just as an OS should not crash when there is a problem with its user applications, so boinc client should not crash when there is a problem with one of its applications.

---

Boinc client has now been running about two months now with no crashes, but getting no Hydrogen@home applications being fetched or run. Pretty much points the finger at H@H.

---