They had database trouble today, and are fixing all the erroneous validate errors even as we speak.
Best thing to do is resume the work you have onboard and just let it run. Last time I checked most of the backend processes were still disabled, so you may run out of work temporarily, but it should all take care of itself once they get everything straightened out again.
Here is a quick summary of what happened in the past 8 hours:
An admin mistake (SQL command update result set outcome=6;validate_state=2 where id=84114386;) accidentally set all the results in the database into an outcome=validate error state (the first semicolon in the command should be a comma!).
I have corrected these as best as I could. There may be a few hundred results which are not quite in the correct state. Please bear with me while I correct these over the next few days.
I have modified the reporting deadlines for any results that were due in the past 8 hours or the next 4 hours, advancing these deadlines by 12 hours. So results will not be marked as late because of this project downtime.
Hopefully my database repairs will be effective and most Einstein@Home contributors should not notice any problems or unusual behavior with the project.
I checked over my account and I don't seem to have any collateral domage to report. Completed, pendings and in progress all seem to be in the correct state.
I've even had one complete and report since the backend came back up (although it had probably been waiting to report for a few hours at least).
LOL... you have to hate those punctuation errors in command lines though!
BTW, if you're going to be in working on database records anyway, I have this task on one of my old timers. It's a reissue from S5R2, but it's one of the long ones and should have never gotten sent to this host at all. However, I have about 480 hours on it and it will complete fine except I need about 2 more weeks to complete it (November 3rd would be fine). That way you don't have to reissue another S5R2 and this old timer can get credit for 5 weeks hard crunchin'! TIA. ;-)
Yes, that was my mistake. This was one of 131 results that I should have left as 'outcome=validation errors' but in my haste I changed this to 'outcome=success'.
10/15/07 21:34:44|Einstein@Home|Restarting task h1_0314.35_S5R2__43_S5R3a_2 using einstein_S5R3 version 407
10/15/07 22:32:25|Einstein@Home|Task h1_0314.35_S5R2__43_S5R3a_2 exited with zero status but no 'finished' file
10/15/07 22:32:25|Einstein@Home|If this happens repeatedly you may need to reset the project.
10/15/07 22:33:13|Einstein@Home|Restarting task h1_0314.35_S5R2__43_S5R3a_2 using einstein_S5R3 version 407
10/15/07 23:13:59||Running CPU benchmarks
10/15/07 23:13:59||Suspending computation - running CPU benchmarks
10/15/07 23:14:31||Benchmark results:
10/15/07 23:14:31|| Number of CPUs: 1
10/15/07 23:14:31|| 1659 floating point MIPS (Whetstone) per CPU
10/15/07 23:14:31|| 3090 integer MIPS (Dhrystone) per CPU
10/15/07 23:14:32||Resuming computation
]
validate errors
)
They had database trouble today, and are fixing all the erroneous validate errors even as we speak.
Best thing to do is resume the work you have onboard and just let it run. Last time I checked most of the backend processes were still disabled, so you may run out of work temporarily, but it should all take care of itself once they get everything straightened out again.
Alinator
Here is a quick summary of
)
Here is a quick summary of what happened in the past 8 hours:
An admin mistake (SQL command update result set outcome=6;validate_state=2 where id=84114386;) accidentally set all the results in the database into an outcome=validate error state (the first semicolon in the command should be a comma!).
I have corrected these as best as I could. There may be a few hundred results which are not quite in the correct state. Please bear with me while I correct these over the next few days.
I have modified the reporting deadlines for any results that were due in the past 8 hours or the next 4 hours, advancing these deadlines by 12 hours. So results will not be marked as late because of this project downtime.
Hopefully my database repairs will be effective and most Einstein@Home contributors should not notice any problems or unusual behavior with the project.
Cheers,
Bruce
Director, Einstein@Home
Well thanks for the update
)
Well thanks for the update Dr. Allen.
I checked over my account and I don't seem to have any collateral domage to report. Completed, pendings and in progress all seem to be in the correct state.
I've even had one complete and report since the backend came back up (although it had probably been waiting to report for a few hours at least).
LOL... you have to hate those punctuation errors in command lines though!
BTW, if you're going to be in working on database records anyway, I have this task on one of my old timers. It's a reissue from S5R2, but it's one of the long ones and should have never gotten sent to this host at all. However, I have about 480 hours on it and it will complete fine except I need about 2 more weeks to complete it (November 3rd would be fine). That way you don't have to reissue another S5R2 and this old timer can get credit for 5 weeks hard crunchin'! TIA. ;-)
Alinator
Is this one of those errors?
)
Is this one of those errors? Notice I got no credit while 2 others have some, but my results seem ok.
http://einsteinathome.org/workunit/34957517
RE: Is this one of those
)
Yes, that was my mistake. This was one of 131 results that I should have left as 'outcome=validation errors' but in my haste I changed this to 'outcome=success'.
I have fixed these 131 results (including yours).
Thanks for pointing it out!
Cheers,
Bruce
Director, Einstein@Home
Is this one of the mistakes?
)
Is this one of the mistakes? http://einsteinathome.org/workunit/34921280
RE: Is this one of the
)
This appears to be a genuine error in the result.
Bruce
Director, Einstein@Home
No "'finished' file"? This
)
No "'finished' file"? This is a first for me--all part of the error? Bits of the log file follow:
10/15/07 12:56:41||Starting BOINC client version 5.10.7 for windows_intelx86
10/15/07 12:56:41||log flags: task, file_xfer, sched_ops
10/15/07 12:56:41||Libraries: libcurl/7.16.1 OpenSSL/0.9.8e zlib/1.2.3
10/15/07 12:56:41||Data directory: C:\\Program Files\\BOINC
10/15/07 12:56:58||Processor: 2 GenuineIntel Intel(R) Core(TM)2 CPU T5600 @ 1.83GHz [x86 Family 6 Model 15 Stepping 6]
10/15/07 12:56:58||Processor features: fpu tsc pae nx sse sse2 mmx
10/15/07 12:56:58||Memory: 2.00 GB physical, 3.85 GB virtual
10/15/07 12:56:58||Disk: 79.17 GB total, 54.34 GB free
10/15/07 12:56:58|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 882874; location: work; project prefs: work
10/15/07 21:34:44|Einstein@Home|Restarting task h1_0314.35_S5R2__43_S5R3a_2 using einstein_S5R3 version 407
10/15/07 22:32:25|Einstein@Home|Task h1_0314.35_S5R2__43_S5R3a_2 exited with zero status but no 'finished' file
10/15/07 22:32:25|Einstein@Home|If this happens repeatedly you may need to reset the project.
10/15/07 22:33:13|Einstein@Home|Restarting task h1_0314.35_S5R2__43_S5R3a_2 using einstein_S5R3 version 407
10/15/07 23:13:59||Running CPU benchmarks
10/15/07 23:13:59||Suspending computation - running CPU benchmarks
10/15/07 23:14:31||Benchmark results:
10/15/07 23:14:31|| Number of CPUs: 1
10/15/07 23:14:31|| 1659 floating point MIPS (Whetstone) per CPU
10/15/07 23:14:31|| 3090 integer MIPS (Dhrystone) per CPU
10/15/07 23:14:32||Resuming computation
Jonathan
RE: RE: Is this one of
)
Actually this looks like a bug in the 4.07 App, probably related to the "new checkpointing code", so the 4.09 might have it, too.
BM
BM
RE: RE: RE: Is this one
)
Then I am glad I brought it up. Something more for you guys to work on.
Thanks for both your updated, Dr. Allen and Bernd (are you a Dr. also?).