Rechenkraft.net e.V.

New in Cruncher 1.2* (Windows, Mac PPC/Intel, Linux32/64, Solaris):
*other operating system will follow*

Now the used cpu time is stored together with the checkpoints. After restart the cpu time will not fall back (at least no more than 20min).

yoyo

PS: Thanks to cody, scsimodo and dotsch.

I am running Cruncher 1.22 on Mac OS X/Intel and the running time with this application is increasing very slowly, altough it is using over 90% of processor time. Is this a known bug or it is a problem with my machine?

Thanks in advance.

Hello,
yes this is a know problem, which is in the nature of the wrapper approach. This CPU throttling is not well supported. Boinc makes it in the way, that the application is stopped for 1 second and resumed afterwards for some seconds and than again stops the application again for 1 second. This is done in a way to reach e.g. 90% CPU usage.
The problem now of the wrapper and the wrapped dnet appliaction is, that it does not react so fast to this. So if boinc stops the application, the dnet app needs some seconds to do it.
This leads to the slow progress.
yoyo

I have a org 1.20 Wu thats been running on my PS3 for close to 60 hr's now, at some time I have to move on and stop wasting the PS3 resources so when do I abort this aberation so I can get on with my PS3's Life & run some Wu's from other Projects that will finish ... It shows no % progress in the Progress Tab either but that may be normal ... ???

PoorBoy hat geschrieben:I have a org 1.20 Wu thats been running on my PS3 for close to 60 hr's now, at some time I have to move on and stop wasting the PS3 resources so when do I abort this aberation so I can get on with my PS3's Life & run some Wu's from other Projects that will finish ... It shows no % progress in the Progress Tab either but that may be normal ... ???

60 hr's is way to long for one ogr WU on a PS3 (My longest WU took about 11 hr's). That the progress bar is showing nothing is normal.
A yoyo WU taking so long and never getting finished is a problem I encountered once when I was running a second project on BOINC (PS3GRID). I figured out, that my WU's never finished when the projects switched while the WU of the project wasn't finished!

I've Suspended the Wu for now, it's @ 66 hr's so far with no idea when or if it will or would ever finish. I doubt I'll ever run anymore Yoyo Wu's on it as thats way to much time to waste on a Wu without a Progress Bar to go by to see it the Wu is Progressing. I could Download another 1 & the same thing could happen so I'm not taking the chance of wasting another 2 or 3 day's of Processing time.

Do you have cpu throttling switched on in Boinc?
Can you send me the corresponding ogr_*_1 file for this wu?
yoyo

No I don't have CPU Throttling enabled Yoyo, below is all that's in the ogr_*_1 file for this wu ... It looks like about 1:00PM on the 17'th was the last real activity for the Wu other than to add CPU Time since then, should or can I abort the Wu now ... ???

distributed.net client for Linux Copyright 1997-2007, distributed.net
RC5-72 Altivec and OGR assembly by Didier Levet
RC5-72 and OGR SPE assembly by Decio Luiz Gazzoni Filho
Please visit http://www.distributed.net/ for up-to-date contest information.

Client will run without network access.
Setting distributed.net ID to u_PoorBoy@yoyo.rechenkraft.net
Client will exit when buffers are empty.
Setting checkpoint file to 'chkpoint'
Setting pause file to 'pause'
Setting exit file to 'exit'
Setting in-buffer base name to in
Setting out-buffer base name to out

dnetc v2.9015-505-CFR-08010411 for Linux (Linux 2.6.23-9.ydl6.1).
Please provide the *entire* version descriptor when submitting bug reports.
The distributed.net bug report pages are at http://www.distributed.net/bugs/
Using email address (distributed.net ID) 'u_PoorBoy@yoyo.rechenkraft.net'

[Sep 17 09:24:25 UTC] Automatic processor detection found 7 processors.
[Sep 17 09:24:25 UTC] Loading crunchers with work...
[Sep 17 09:24:25 UTC] Automatic processor type detection found
a Cell Broadband Engine processor.
[Sep 17 09:24:25 UTC] OGR-P2: using core #1 (KOGE 2.0 Hybrid).
[Sep 17 09:24:25 UTC] OGR-P2 #a: Loaded 25/13-9-24-60
[Sep 17 09:24:25 UTC] OGR-P2 #b: Loaded 25/13-9-24-61
[Sep 17 09:24:25 UTC] OGR-P2 #c: Loaded 25/13-9-24-62
[Sep 17 09:24:25 UTC] OGR-P2 #d: Loaded 25/13-9-24-63
[Sep 17 09:24:25 UTC] OGR-P2 #e: Loaded 25/13-9-24-65
[Sep 17 09:24:25 UTC] OGR-P2 #f: Loaded 25/13-9-24-66
[Sep 17 09:24:25 UTC] OGR-P2 #g: Loaded 25/13-9-24-67
[Sep 17 09:24:25 UTC] OGR-P2: 2 packets remain in in.ogf
[Sep 17 09:24:25 UTC] OGR-P2: 0 packets are in out.ogf
[Sep 17 09:24:25 UTC] 7 crunchers ('a'-'g') have been started.

[Sep 17 09:25:00 UTC] OGR-P2 #d: Completed 25/13-9-24-63 (0.80 stats units)
0.00:00:34.52 - [23,296,577 nodes/s]
[Sep 17 09:25:00 UTC] OGR-P2 #d: 25/13-9-24-63 [804,197,864 nodes]
[Sep 17 09:25:00 UTC] OGR-P2 #d: Loaded 25/13-9-24-68
[Sep 17 09:25:00 UTC] OGR-P2: Summary: 1 packet (0.80 stats units)
0.00:00:34.52 - [23.30 Mnodes/s]
[Sep 17 09:25:00 UTC] OGR-P2: 1 packet remains in in.ogf
[Sep 17 09:25:00 UTC] OGR-P2: 1 packet (0.80 stats units) is in out.ogf

[Sep 17 11:51:05 UTC] OGR-P2 #g: Completed 25/13-9-24-67 (142.53 stats units)
0.02:26:38.98 - [16,198,600 nodes/s]
[Sep 17 11:51:05 UTC] OGR-P2 #g: 25/13-9-24-67 [142,531,258,117 nodes]
[Sep 17 11:51:05 UTC] OGR-P2 #g: Loaded 25/13-9-24-69
[Sep 17 11:51:05 UTC] OGR-P2: Summary: 2 packets (143.33 stats units)
0.02:26:39.70 - [16.29 Mnodes/s]
[Sep 17 11:51:05 UTC] OGR-P2: 0 packets remain in in.ogf
[Sep 17 11:51:05 UTC] OGR-P2: 2 packets (143.33 stats units) are in out.ogf

[Sep 17 13:00:57 UTC] OGR-P2 #a: Completed 25/13-9-24-60 (228.14 stats units)
0.03:36:30.74 - [17,561,391 nodes/s]
[Sep 17 13:00:57 UTC] OGR-P2 #a: 25/13-9-24-60 [228,135,575,957 nodes]
[Sep 17 13:00:57 UTC] OGR-P2: Summary: 3 packets (371.47 stats units)
0.03:36:31.66 - [28.59 Mnodes/s]
[Sep 17 13:00:57 UTC] OGR-P2: 0 packets remain in in.ogf
[Sep 17 13:00:57 UTC] OGR-P2: 3 packets (371.47 stats units) are in out.ogf

Yes, it completed 3 ogr but not the remaining 6.
Can you stop/restart boinc so this wu gets restarted?
Afterwards something new in the ogr_*_1 should appear. If not you should abort the WU.
yoyo

The org_1 File looks like this now after re-starting BOINC, it looks like theres some activity again with the Wu, I'll let it run some more & see what happens. How often should there be some activity with this file so I can keep an eye on it & see if the Wu is progressing or not. I'd hate to spend another 66 hours on it again ...

distributed.net client for Linux Copyright 1997-2007, distributed.net
RC5-72 Altivec and OGR assembly by Didier Levet
RC5-72 and OGR SPE assembly by Decio Luiz Gazzoni Filho
Please visit http://www.distributed.net/ for up-to-date contest information.

Client will run without network access.
Setting distributed.net ID to u_PoorBoy@yoyo.rechenkraft.net
Client will exit when buffers are empty.
Setting checkpoint file to 'chkpoint'
Setting pause file to 'pause'
Setting exit file to 'exit'
Setting in-buffer base name to in
Setting out-buffer base name to out

dnetc v2.9015-505-CFR-08010411 for Linux (Linux 2.6.23-9.ydl6.1).
Please provide the *entire* version descriptor when submitting bug reports.
The distributed.net bug report pages are at http://www.distributed.net/bugs/
Using email address (distributed.net ID) 'u_PoorBoy@yoyo.rechenkraft.net'

[Sep 17 09:24:25 UTC] Automatic processor detection found 7 processors.
[Sep 17 09:24:25 UTC] Loading crunchers with work...
[Sep 17 09:24:25 UTC] Automatic processor type detection found
a Cell Broadband Engine processor.
[Sep 17 09:24:25 UTC] OGR-P2: using core #1 (KOGE 2.0 Hybrid).
[Sep 17 09:24:25 UTC] OGR-P2 #a: Loaded 25/13-9-24-60
[Sep 17 09:24:25 UTC] OGR-P2 #b: Loaded 25/13-9-24-61
[Sep 17 09:24:25 UTC] OGR-P2 #c: Loaded 25/13-9-24-62
[Sep 17 09:24:25 UTC] OGR-P2 #d: Loaded 25/13-9-24-63
[Sep 17 09:24:25 UTC] OGR-P2 #e: Loaded 25/13-9-24-65
[Sep 17 09:24:25 UTC] OGR-P2 #f: Loaded 25/13-9-24-66
[Sep 17 09:24:25 UTC] OGR-P2 #g: Loaded 25/13-9-24-67
[Sep 17 09:24:25 UTC] OGR-P2: 2 packets remain in in.ogf
[Sep 17 09:24:25 UTC] OGR-P2: 0 packets are in out.ogf
[Sep 17 09:24:25 UTC] 7 crunchers ('a'-'g') have been started.

[Sep 17 09:25:00 UTC] OGR-P2 #d: Completed 25/13-9-24-63 (0.80 stats units)
0.00:00:34.52 - [23,296,577 nodes/s]
[Sep 17 09:25:00 UTC] OGR-P2 #d: 25/13-9-24-63 [804,197,864 nodes]
[Sep 17 09:25:00 UTC] OGR-P2 #d: Loaded 25/13-9-24-68
[Sep 17 09:25:00 UTC] OGR-P2: Summary: 1 packet (0.80 stats units)
0.00:00:34.52 - [23.30 Mnodes/s]
[Sep 17 09:25:00 UTC] OGR-P2: 1 packet remains in in.ogf
[Sep 17 09:25:00 UTC] OGR-P2: 1 packet (0.80 stats units) is in out.ogf

[Sep 17 11:51:05 UTC] OGR-P2 #g: Completed 25/13-9-24-67 (142.53 stats units)
0.02:26:38.98 - [16,198,600 nodes/s]
[Sep 17 11:51:05 UTC] OGR-P2 #g: 25/13-9-24-67 [142,531,258,117 nodes]
[Sep 17 11:51:05 UTC] OGR-P2 #g: Loaded 25/13-9-24-69
[Sep 17 11:51:05 UTC] OGR-P2: Summary: 2 packets (143.33 stats units)
0.02:26:39.70 - [16.29 Mnodes/s]
[Sep 17 11:51:05 UTC] OGR-P2: 0 packets remain in in.ogf
[Sep 17 11:51:05 UTC] OGR-P2: 2 packets (143.33 stats units) are in out.ogf

[Sep 17 13:00:57 UTC] OGR-P2 #a: Completed 25/13-9-24-60 (228.14 stats units)
0.03:36:30.74 - [17,561,391 nodes/s]
[Sep 17 13:00:57 UTC] OGR-P2 #a: 25/13-9-24-60 [228,135,575,957 nodes]
[Sep 17 13:00:57 UTC] OGR-P2: Summary: 3 packets (371.47 stats units)
0.03:36:31.66 - [28.59 Mnodes/s]
[Sep 17 13:00:57 UTC] OGR-P2: 0 packets remain in in.ogf
[Sep 17 13:00:57 UTC] OGR-P2: 3 packets (371.47 stats units) are in out.ogf

distributed.net client for Linux Copyright 1997-2007, distributed.net
RC5-72 Altivec and OGR assembly by Didier Levet
RC5-72 and OGR SPE assembly by Decio Luiz Gazzoni Filho
Please visit http://www.distributed.net/ for up-to-date contest information.

Client will run without network access.
Setting distributed.net ID to u_PoorBoy@yoyo.rechenkraft.net
Client will exit when buffers are empty.
Setting checkpoint file to 'chkpoint'
Setting pause file to 'pause'
Setting exit file to 'exit'
Setting in-buffer base name to in
Setting out-buffer base name to out

dnetc v2.9015-505-CFR-08010411 for Linux (Linux 2.6.23-9.ydl6.1).
Please provide the *entire* version descriptor when submitting bug reports.
The distributed.net bug report pages are at http://www.distributed.net/bugs/
Using email address (distributed.net ID) 'u_PoorBoy@yoyo.rechenkraft.net'

[Sep 20 10:15:13 UTC] Recovered 6 checkpoint packets
[Sep 20 10:15:13 UTC] Automatic processor detection found 7 processors.
[Sep 20 10:15:13 UTC] Loading crunchers with work...
[Sep 20 10:15:13 UTC] Automatic processor type detection found
a Cell Broadband Engine processor.
[Sep 20 10:15:13 UTC] OGR-P2: using core #1 (KOGE 2.0 Hybrid).
[Sep 20 10:15:13 UTC] OGR-P2 #a: Loaded 25/13-9-24-61
Packet was from a different user/core/client cpu/os/b ...
[Sep 20 10:15:13 UTC] OGR-P2 #b: Loaded 25/13-9-24-62 (1.66 Tnodes done)
[Sep 20 10:15:13 UTC] OGR-P2 load failure: CORE_E_STUB: Invalid initial ruler
Stub discarded.
[Sep 20 10:15:13 UTC] OGR-P2 #c: Loaded 25/13-9-24-65 (4.33 Tnodes done)
[Sep 20 10:15:13 UTC] OGR-P2 load failure: CORE_E_STUB: Invalid initial ruler
Stub discarded.
[Sep 20 10:15:13 UTC] OGR-P2 #d: Loaded 25/13-9-24-69 (2.59 Tnodes done)
[Sep 20 10:15:13 UTC] OGR-P2: 0 packets remain in in.ogf
[Sep 20 10:15:13 UTC] OGR-P2: 3 packets (371.47 stats units) are in out.ogf
[Sep 20 10:15:13 UTC] 4 crunchers ('a'-'d') have been started.

Hello,
this file is only updated, if one of the contained ogr are finished. I wonder a bit, because the client says that 6 checkpoints are recovered, but only 4 crunchers (a-d) are started. Is there a second project in boinc, which runs also?
You can see progress also on the checkpoint file "chkpoint" in slot directory, which should update nearly every 15 minutes.
yoyo

Yoyo, this is a PlayStation3 thats running this Wu, as far as I know only 1 Wu will run @ a time on the PS3's. There are other Projects Attached to the PS3 but the Wu's aren't running, only the Yoyo Wu is running as far as I know. The ckpoint file won't open up into anything thats readable by me ... ???