CC4.16 killed one of my clients
22 Jan 2005 3:25:54 UTC
Topic 187291
(moderation:
On my Win98se machine, on both NT4 machines and on on of my W2K machines the CC4.16 upgrade caused nio problems what so ever.
On one W2K/SP3 machine, it hosed things but good - Among other things it told me that the LHC, PP@H and Pirates Host IDs weren't assigned. It also killed the inflight CPDN WU. All I had left were two S@H WUs.
v4.15
2005-01-21 18:55:33 [---] Starting BOINC client version 4.15 for windows_intelx86
2005-01-21 18:55:33 [climateprediction.net] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:33 [Einstein@Home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:33 [LHC@home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:33 [Pirates@Home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:33 [ProteinPredictorAtHome] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:33 [SETI@home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:33 [climateprediction.net] Host ID is 82092
2005-01-21 18:55:33 [Einstein@Home] Host ID is 5006
2005-01-21 18:55:33 [LHC@home] Host ID is 6255
2005-01-21 18:55:33 [Pirates@Home] Host ID is 1403
2005-01-21 18:55:33 [ProteinPredictorAtHome] Host ID is 30361
2005-01-21 18:55:33 [SETI@home] Host ID is 406340
2005-01-21 18:55:33 [---] General prefs: from SETI@home (last modified 2005-01-20 15:18:21)
2005-01-21 18:55:33 [---] General prefs: using separate prefs for home
2005-01-21 18:55:38 [climateprediction.net] Deferring computation for result 2tgh_000152816_4
2005-01-21 18:55:38 [Einstein@Home] Resuming computation for result H1_0205.4__0205.6_0.1_T08_Test02_4 using einstein version 4.71
2005-01-21 18:55:38 [ProteinPredictorAtHome] Deferring computation for result t0243D_1_3710_0
2005-01-21 18:55:38 [SETI@home] Deferring computation for result 27mr04aa.17618.10704.322166.211_0
2005-01-21 18:55:38 [---] May run out of work in 3.00 days; requesting more
2005-01-21 18:55:38 [SETI@home] Requesting 19072 seconds of work
2005-01-21 18:55:38 [SETI@home] Sending request to scheduler: http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
2005-01-21 18:55:41 [SETI@home] Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
2005-01-21 18:55:41 [SETI@home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:55:41 [SETI@home] Started download of 03ja04aa.13491.15073.598592.230
v4.16
2005-01-21 18:47:36 [---] Starting BOINC client version 4.16 for windows_intelx86
2005-01-21 18:47:36 [climateprediction.net] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:47:36 [Einstein@Home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:47:36 [LHC@home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:47:36 [Pirates@Home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:47:36 [ProteinPredictorAtHome] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:47:36 [SETI@home] Project prefs: no separate prefs for home; using your defaults
2005-01-21 18:47:36 [---] Version Change Detected (4.15 -> 4.16); running CPU benchmarks
2005-01-21 18:47:36 [climateprediction.net] Host ID is 82092
2005-01-21 18:47:36 [Einstein@Home] Host ID is 5006
2005-01-21 18:47:36 [LHC@home] Host ID not assigned yet
2005-01-21 18:47:36 [Pirates@Home] Host ID not assigned yet
2005-01-21 18:47:36 [ProteinPredictorAtHome] Host ID not assigned yet
2005-01-21 18:47:36 [SETI@home] Host ID is 406340
2005-01-21 18:47:36 [---] General prefs: from SETI@home (last modified 2005-01-20 15:18:21)
2005-01-21 18:47:36 [---] General prefs: using separate prefs for home
2005-01-21 18:47:41 [---] Running CPU benchmarks
Luckily I learned from the 4.09 -> 4.12 fiasco and backed up the entire BOINC directory before upgrading.
Any thoughts, anyone (and no, I have had no discernable problems with S@H/PP@H/CPDN on this machine with 4.15. The machine is happily crunching away again on 4.15)
CC4.16 killed one of my clients
)
Same problem with me. On my WinXP Home SP2, I lost 800 hours worth of calculation for climate prediction (I was only 24 hours from returning the results !!!).
As it's the first time I had such problem, I did not backed up my Boinc directory !
Shit happens, I guess (pardon my french)
xipehuz
I also lost some WUs from
)
I also lost some WUs from 4.16. It killed Predictor and Einstein, but left S@H and CPDN alone. I did roll back to 4.15 and restored a backup because I thought 4.16 was affecting the scheduler on Predictor, but the scheduler was just down, so I went back to 4.16. No problems running it to report.
Don't worry about the message saying a host ID isn't assigned. As soon as the scheduler is contacted on the project, it will give a new host ID. This is because the new CC reads the CPU name differently. (More accurately. No more AMD Pentiums...) The projects that give this message will now have another host listed under your computers that you may not be able to merge, depending on how the CPU is now defined.
> I also lost some WUs from
)
> I also lost some WUs from 4.16. It killed Predictor and Einstein, but left S@H
> and CPDN alone. I did roll back to 4.15 and restored a backup because I
> thought 4.16 was affecting the scheduler on Predictor, but the scheduler was
> just down, so I went back to 4.16. No problems running it to report.
>
> Don't worry about the message saying a host ID isn't assigned. As soon as the
> scheduler is contacted on the project, it will give a new host ID. This is
> because the new CC reads the CPU name differently. (More accurately. No more
> AMD Pentiums...) The projects that give this message will now have another
> host listed under your computers that you may not be able to merge, depending
> on how the CPU is now defined.
>
>
OK, I can accept the explaination about the Project ID, but out of Seti, Predictor, Einstein and CPDN WUs in process on my machine, only the Seti WUs were recognized and retained . After the upgrade, all the others were gone. That's several hundred CPU hours for the CPDN WU.
OK, I can accept the
)
OK, I can accept the explaination about the Project ID, but out of Seti, Predictor, Einstein and CPDN WUs in process on my machine, only the Seti WUs were recognized and retained . After the upgrade, all the others were gone. That's several hundred CPU hours for the CPDN WU.
=========
A lesson learned I hope, Backup, Backup, Backup is all I can say whenever you do a Version Upgrade. I had the same thing happen to me on 1 Computer yesterday because I was to lazy to do a Backup first, I knew I should have when I did the Upgrade but I figured it went smooth on another Computer so I didn't bother Backing up the second computer first.
The rest of my Computers I saved the BOINC Folder some place else first & renamed it, a couple of them didn't take the Upgrade either but I just copied the Saved Folder back to the original directory and just started out where I left off from ... :)
> OK, I can accept the
)
> OK, I can accept the explaination about the Project ID, but out of Seti,
> Predictor, Einstein and CPDN WUs in process on my machine, only the Seti WUs
> were recognized and retained . After the upgrade, all the others were gone.
> That's several hundred CPU hours for the CPDN WU.
> =========
>
> A lesson learned I hope, Backup, Backup, Backup is all I can say whenever you
> do a Version Upgrade. I had the same thing happen to me on 1 Computer
> yesterday because I was to lazy to do a Backup first, I knew I should have
> when I did the Upgrade but I figured it went smooth on another Computer so I
> didn't bother Backing up the second computer first.
>
> The rest of my Computers I saved the BOINC Folder some place else first &
> renamed it, a couple of them didn't take the Upgrade either but I just copied
> the Saved Folder back to the original directory and just started out where I
> left off from ... :)
>
>
yeah - I did. I did in fact learn from the 4.01 ->4.12 fiasco. My phrasing is a little unclear. After restoring the backup of 4.15, S@H,PP@H, and E@H have all returned WUs, and CPDN is happily time-slicing and crunching.