I figured I should post the "unrecoverable error" below just in case you wanted to know about it; however everything is working/proceeding fine. It seems to have occurred when boinc was switching to a second project (wcg).

Host Project Date Message
Tinkerbell Tinkerbell rosetta@home 11/12/2005 4:03:26 PM Finished upload of 1hz7A_abrelaxmode_random_gauss_fix_bb_jitter03_130292_0_0
Tinkerbell Tinkerbell rosetta@home 11/12/2005 4:03:26 PM Throughput 33096 bytes/sec
Tinkerbell Tinkerbell World Community Grid 11/12/2005 4:43:20 PM Restarting result de147_51_1 using rosetta version 419
Tinkerbell Tinkerbell rosetta@home 11/12/2005 4:43:20 PM Pausing result 1hz7A_abrelaxmode_random_gauss_fix_bb_jitter03_130304_0 (removed from memory)
Tinkerbell Tinkerbell rosetta@home 11/12/2005 4:43:21 PM Unrecoverable error for result 1hz7A_abrelaxmode_random_gauss_fix_bb_jitter03_130304_0 (process got signal 11)
Tinkerbell Tinkerbell --- 11/12/2005 4:43:21 PM request_reschedule_cpus: process exited
Tinkerbell Tinkerbell rosetta@home 11/12/2005 4:43:21 PM Deferring communication with project for 1 minutes and 0 seconds
Tinkerbell Tinkerbell rosetta@home 11/12/2005 4:43:21 PM Computation for result 1hz7A_abrelaxmode_random_gauss_fix_bb_jitter03_130304_0 finished
Tinkerbell Tinkerbell World Community Grid 11/12/2005 5:03:22 PM Pausing result de147_51_1 (removed from memory)
Tinkerbell Tinkerbell rosetta@home 11/12/2005 5:03:22 PM Starting result 1hz7A_abrelaxmode_random_gauss_fix_bb_jitter03_130263_0 using rosetta version 479

I figured I should post the "unrecoverable error" below just in case you wanted to know about it; however everything is working/proceeding fine. It seems to have occurred when boinc was switching to a second project (wcg).

From the logs you posted, it looks like you're removing the app from memory on switching. There is a known problem with this for Rosetta, as noted in several other threads. Suggest you set your "Leave applications in memory while preempted?" preference to YES and see if the problem disappears.

---

*** Join BOINC@Australia today ***

Thanks for the tip Yoda. This error isnt bothering me yet and when I previously had the 'leave in memory' enabled I got messages from WCG that there wasnt enough memory, so it seemed important to provide the most memory possible to projects by letting them preempt out of memory -- even if I have to bump them along if they error out. However if this gets unmanageable I'll make the preference to stay in memory but remove the WCG project for now.

---

If you can't leave it in memory, you could increase your "switch every" value to something longer than it takes for you to complete a Rosetta WU. (180 minutes should be WAYYY more than enough)

---

Questions? Try the Wiki!
Take a look inside my brain

(180 minutes should be WAYYY more than enough)

Depends on the CPU speed and work unit. Slowest WUs on my PC's in the last 24 hours (all WUs of the 1hz6A variety):

3.4GHz Pentium 4 with HT: 4 hours
2.8GHz Pentium 4 no HT: 3.75 hours
2.4GHz Pentium 4 no HT: 5 hours
Athlon 64 3700+ (at 2.64GHz): 2 hours
Athlon XP 3000+ (at 2.35Ghz): 3.3 hours

With what you (eberndl) recommended, 4 of these 5 work units (and others) might have crashed.

---

*** Join BOINC@Australia today ***