S@H Enhanced v6.04

Message boards : Number crunching : S@H Enhanced v6.04
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1149933 - Posted: 8 Sep 2011, 2:36:58 UTC
Last modified: 8 Sep 2011, 2:37:32 UTC

Got home from work today and found this Task on one of my G4s. It is flagged as v6.04, with a hugely over-long estimated completion time of 166 hrs (vlars normally take this box about 40 hrs), and a new entry in the Details section for S@H v6.04.

Is this a new ap version, or a glitch? If a new ap, what changed?
Donald
Infernal Optimist / Submariner, retired
ID: 1149933 · Report as offensive
Profile Dirk Sadowski
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 1149940 - Posted: 8 Sep 2011, 2:50:20 UTC
Last modified: 8 Sep 2011, 2:55:04 UTC

It looks like it's a new app for Mac OS/X 10.3+ .
http://setiathome.berkeley.edu/apps.php

AFAIK, after 10 granted Cr./WUs with this new app and the estimated times should be again correct.


But little strange..
If this overview is correct, this app wasn't testet at SETI@home Beta Test.
http://setiweb.ssl.berkeley.edu/beta/apps.php

So maybe it's a mistake.

Or maybe only a cosmetic modification of the app.


- Best regards! - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC. - SETI@home needs your help. -
ID: 1149940 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1149946 - Posted: 8 Sep 2011, 3:05:04 UTC - in response to Message 1149940.  

It looks like it's a new app for Mac OS/X 10.3+ .
http://setiathome.berkeley.edu/apps.php

AFAIK, after 10 granted Cr./WUs with this new app and the estimated times should be again correct.

But little strange..
If this overview is correct, this app wasn't testet at SETI@home Beta Test.
http://setiweb.ssl.berkeley.edu/beta/apps.php

So maybe it's a mistake.

Or maybe only a cosmetic modification of the app.

Oh, joy! That means my other G4 will get its first 6.04 on Thursday. 10 tasks will take about 2 weeks for either box, unless we get another shorty storm.

Well, at least the current estimates are too long, rather than too short. Plus, each validated task should start scaling the DCF back to where it belongs. Just gonna be fun to watch for the next two weeks.

Donald
Infernal Optimist / Submariner, retired
ID: 1149946 · Report as offensive
Profile Dirk Sadowski
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 1149949 - Posted: 8 Sep 2011, 3:15:03 UTC - in response to Message 1149946.  
Last modified: 8 Sep 2011, 3:41:29 UTC

Ohh.. or was this 10 granted Cr./WUs and the following Cr./WU are correct?
Yes I guess this is correct.

Because of the estimated times it will go down slowly, independent of granted Cr./WUs.
Every time a new result, the estimated times of the WUs go little down.

So maybe it will last some days/weeks if the estimated times are again correct.

Sorry, my mistake.


- Best regards! - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC. - SETI@home needs your help. -
ID: 1149949 · Report as offensive
Profile Dirk Sadowski
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 1149963 - Posted: 8 Sep 2011, 3:40:51 UTC
Last modified: 8 Sep 2011, 3:54:29 UTC

BTW.

You are an advanced SETI@home member?

You know there are also optimized applications for MACs available?

(Message 1112592)

Hmm.. maybe currently they are outdated.
If stock 6.04 is really different compared to 6.03, someone will release quickly also updated optimized apps.


- Best regards! - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC. - SETI@home needs your help. -
ID: 1149963 · Report as offensive
Profile arkayn
Volunteer tester
Avatar

Send message
Joined: 14 May 99
Posts: 4438
Credit: 55,006,323
RAC: 0
United States
Message 1149970 - Posted: 8 Sep 2011, 4:09:05 UTC

Must only be a minor update as this is the first we have heard of it.

ID: 1149970 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1149979 - Posted: 8 Sep 2011, 5:38:11 UTC - in response to Message 1149970.  

Must only be a minor update as this is the first we have heard of it.

There were no new commits to the SETI@home Enhanced source code so I think it's probably just a rebuild using a more recent compiler than the 6.03 version.

They did try to build a 6.97 SETI@home v7 Mac application for SETI Beta, but the currently supported Apple version of GCC doesn't have support for AVX code so they left the earlier 6.92 in place there. Last time I looked, Apple had GCC 4.2.1 but at least 4.4 is needed for AVX. It's a puzzle that Apple is producing machines with Sandy Bridge CPUs but doesn't support the new instructions yet. That would only apply to Mac/Intel machines of course.
                                                                   Joe
ID: 1149979 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19369
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1149980 - Posted: 8 Sep 2011, 5:42:48 UTC - in response to Message 1149946.  

Oh, joy! That means my other G4 will get its first 6.04 on Thursday. 10 tasks will take about 2 weeks for either box, unless we get another shorty storm.

Well, at least the current estimates are too long, rather than too short. Plus, each validated task should start scaling the DCF back to where it belongs. Just gonna be fun to watch for the next two weeks.


Beware of the bolded text, as you said the client side dcf will go low, maybe v. low, before 10 tasks are validated and the server side estimates are corrected. At that point the server will possibly give the correct estimate, but the local dcf will then lower those estimates. The net result could be d/load of hundreds of tasks.
ID: 1149980 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1149983 - Posted: 8 Sep 2011, 6:09:26 UTC - in response to Message 1149980.  

Oh, joy! That means my other G4 will get its first 6.04 on Thursday. 10 tasks will take about 2 weeks for either box, unless we get another shorty storm.

Well, at least the current estimates are too long, rather than too short. Plus, each validated task should start scaling the DCF back to where it belongs. Just gonna be fun to watch for the next two weeks.


Beware of the bolded text, as you said the client side dcf will go low, maybe v. low, before 10 tasks are validated and the server side estimates are corrected. At that point the server will possibly give the correct estimate, but the local dcf will then lower those estimates. The net result could be d/load of hundreds of tasks.

Yes, that is one thing I'll be watching for, but I doubt I have to worry about hundreds of Tasks with too-short estimates. These are 10-year-old mac G4s, not relatively new i-5s, and my caches are set to 0+4. Maybe I'll change that to 0+2 until things settle out.
Donald
Infernal Optimist / Submariner, retired
ID: 1149983 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19369
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1149994 - Posted: 8 Sep 2011, 7:19:46 UTC - in response to Message 1149983.  

Oh, joy! That means my other G4 will get its first 6.04 on Thursday. 10 tasks will take about 2 weeks for either box, unless we get another shorty storm.

Well, at least the current estimates are too long, rather than too short. Plus, each validated task should start scaling the DCF back to where it belongs. Just gonna be fun to watch for the next two weeks.


Beware of the bolded text, as you said the client side dcf will go low, maybe v. low, before 10 tasks are validated and the server side estimates are corrected. At that point the server will possibly give the correct estimate, but the local dcf will then lower those estimates. The net result could be d/load of hundreds of tasks.

Yes, that is one thing I'll be watching for, but I doubt I have to worry about hundreds of Tasks with too-short estimates. These are 10-year-old mac G4s, not relatively new i-5s, and my caches are set to 0+4. Maybe I'll change that to 0+2 until things settle out.

Don't be too sure, on my quad which I recently had to rebuild, I also installed Win7 at same time, the client DCF went down to 0.087 just before server side estimate for GPU tasks kicked in. Some of the DCF reduction was because Seti servers were off-line for a few days at the critical period.
ID: 1149994 · Report as offensive
Profile arkayn
Volunteer tester
Avatar

Send message
Joined: 14 May 99
Posts: 4438
Credit: 55,006,323
RAC: 0
United States
Message 1150237 - Posted: 8 Sep 2011, 22:37:19 UTC - in response to Message 1149979.  

Must only be a minor update as this is the first we have heard of it.

There were no new commits to the SETI@home Enhanced source code so I think it's probably just a rebuild using a more recent compiler than the 6.03 version.

They did try to build a 6.97 SETI@home v7 Mac application for SETI Beta, but the currently supported Apple version of GCC doesn't have support for AVX code so they left the earlier 6.92 in place there. Last time I looked, Apple had GCC 4.2.1 but at least 4.4 is needed for AVX. It's a puzzle that Apple is producing machines with Sandy Bridge CPUs but doesn't support the new instructions yet. That would only apply to Mac/Intel machines of course.
                                                                   Joe


Looks like Apple is not going to support the newer builds either as they have changed to clang as the supported environment.

Charlie might want to look into Macports instead. They have GCC up to 4.6.1 currently.

ID: 1150237 · Report as offensive
Profile Dirk Sadowski
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 1150613 - Posted: 9 Sep 2011, 23:27:25 UTC
Last modified: 9 Sep 2011, 23:51:37 UTC

S@h v6.04 is little bit buggy?

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4421127
http://setiathome.berkeley.edu/show_host_detail.php?hostid=4614705
http://setiathome.berkeley.edu/show_host_detail.php?hostid=4711242
http://setiathome.berkeley.edu/show_host_detail.php?hostid=5648196
http://setiathome.berkeley.edu/show_host_detail.php?hostid=6135473

Or what is the problem here?


[EDIT: The app overview say this new 6.04 app is for OS 10.3+, why get lower OSs this new app? They should get 6.03?
Why make the app with OS 10.5 and 11.1 errors?]


- Best regards! - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC. - SETI@home needs your help. -
ID: 1150613 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1150661 - Posted: 10 Sep 2011, 2:16:51 UTC - in response to Message 1150613.  

S@h v6.04 is little bit buggy?

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4421127
http://setiathome.berkeley.edu/show_host_detail.php?hostid=4614705
http://setiathome.berkeley.edu/show_host_detail.php?hostid=4711242
http://setiathome.berkeley.edu/show_host_detail.php?hostid=5648196
http://setiathome.berkeley.edu/show_host_detail.php?hostid=6135473

Or what is the problem here?


[EDIT: The app overview say this new 6.04 app is for OS 10.3+, why get lower OSs this new app? They should get 6.03?
Why make the app with OS 10.5 and 11.1 errors?]


I'm not sure of the issue with the "failed: errno=88". However the oldest OS I see there is Darwin 9.8 which is MAC OS 10.5.8
Darwin 9.x = 10.5.x Leopard
Darwin 10.x = 10.6.x Snow Leopard
Darwin 11.x = 10.7.x Lion

It is sort of like converting 12 hr time and 24 hr time. Darwin version -4 = MAC OS version.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1150661 · Report as offensive
Profile Dirk Sadowski
Volunteer tester

Send message
Joined: 6 Apr 07
Posts: 7105
Credit: 147,663,825
RAC: 5
Germany
Message 1150678 - Posted: 10 Sep 2011, 3:44:25 UTC - in response to Message 1150661.  
Last modified: 10 Sep 2011, 3:59:44 UTC

Ahh.. thanks for translation.. ;-)

I didn't know that Darwin 9.8 mean 10.5 MAC OS.

I thought if there stay 10.3+, then 9.8 is smaller.. ;-D


For example, Donald's (OP) PCs have Darwin 8.11, this is also 10.3+ MAC OS?
[EDIT: If I calculated correct it's 10.4 MAC OS. ;-)]


Hmm.. but I guess there is something with 6.04 if 6 MACs don't like it..
I know one more MAC because of an other thread in Q&A: 'Computation Error - Message 1150480'.


- Best regards! - Sutaru Tsureku, team seti.international founder. - Optimize your PC for higher RAC. - SETI@home needs your help. -
ID: 1150678 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1150680 - Posted: 10 Sep 2011, 3:56:41 UTC - in response to Message 1150661.  

S@h v6.04 is little bit buggy?

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4421127
http://setiathome.berkeley.edu/show_host_detail.php?hostid=4614705
http://setiathome.berkeley.edu/show_host_detail.php?hostid=4711242
http://setiathome.berkeley.edu/show_host_detail.php?hostid=5648196
http://setiathome.berkeley.edu/show_host_detail.php?hostid=6135473

Or what is the problem here?

[EDIT: The app overview say this new 6.04 app is for OS 10.3+, why get lower OSs this new app? They should get 6.03?
Why make the app with OS 10.5 and 11.1 errors?]


I'm not sure of the issue with the "failed: errno=88". However the oldest OS I see there is Darwin 9.8 which is MAC OS 10.5.8
Darwin 9.x = 10.5.x Leopard
Darwin 10.x = 10.6.x Snow Leopard
Darwin 11.x = 10.7.x Lion

It is sort of like converting 12 hr time and 24 hr time. Darwin version -4 = MAC OS version.

Here is another one to look at
http://setiathome.berkeley.edu/workunit.php?wuid=814725863
I have the _2 Task, it will start on Saturday 09 Sep 2011. I am running Darwin 8.11.0 = OS X 10.4.11 (missed the upgrade to 10.5), we'll see how my ancient G4 handles it.

Donald
Infernal Optimist / Submariner, retired
ID: 1150680 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1150699 - Posted: 10 Sep 2011, 4:31:03 UTC

If the new app is buggy I might have to fire up the Mac's at work to help out. Currently I just have these machines in the lab:
Dual G5 Power Mac, with MAC OS 10.3, 10.4, or 10.5
Core Duo iMac, with MAC OS 10.4, 10.5, or 10.6(x86 only)
Core 2 Duo iMac, with MAC OS 10.6 or 10.7
I can run any of those configurations if need be. I don't normally run BOINC on those systems anymore since it was a bit of a hassle with how we load OS's for test.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1150699 · Report as offensive
Sami

Send message
Joined: 12 Aug 99
Posts: 38
Credit: 12,671,175
RAC: 4
Finland
Message 1150720 - Posted: 10 Sep 2011, 5:23:27 UTC - in response to Message 1150678.  

Hmm.. but I guess there is something with 6.04 if 6 MACs don't like it..


My Mac doesn't like it either:

http://setiathome.berkeley.edu/results.php?hostid=4674202
ID: 1150720 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1150856 - Posted: 10 Sep 2011, 15:32:09 UTC
Last modified: 10 Sep 2011, 15:35:16 UTC

Well.

The VLAR http://setiathome.berkeley.edu/workunit.php?wuid=788127643 with the 166 hr est. run time that got this thread started, completed in 36 minutes, with a -9 result_overflow, but it validated against a Wintel cruncher, so I'm not worried.

That box (iBook G4) also completed this 0.74 AR Task http://setiathome.berkeley.edu/workunit.php?wuid=815237216 in about 5 minutes with a -9 result_overflow. We'll see if it validates.
Donald
Infernal Optimist / Submariner, retired
ID: 1150856 · Report as offensive
Tomppa
Volunteer tester

Send message
Joined: 12 May 03
Posts: 1
Credit: 812,789
RAC: 0
Finland
Message 1150863 - Posted: 10 Sep 2011, 15:51:10 UTC
Last modified: 10 Sep 2011, 15:54:27 UTC

Hate to say, but my mac don't like 6.04 either:

http://setiathome.berkeley.edu/results.php?hostid=5865300
ID: 1150863 · Report as offensive
Duc Phuc
Volunteer tester

Send message
Joined: 3 Oct 06
Posts: 3
Credit: 1,468,754
RAC: 0
United States
Message 1150865 - Posted: 10 Sep 2011, 15:55:50 UTC - in response to Message 1150720.  

Ditto here. Both my macs started to generate these errors with 6.0.4. I'm having a horrible time trying to reset the project too now. I have two macs both running the current Lion, 11.1.0 and 11.0.1 on a iMac i7 and a mac mini 2011 server i7, respectively.
http://setiathome.berkeley.edu/show_host_detail.php?hostid=6116582
http://setiathome.berkeley.edu/show_host_detail.php?hostid=6118978
I currently suspended the project on those two machines. I'm going to try the optimized application route for these machines. I just did that for my Windows machines.
ID: 1150865 · Report as offensive
1 · 2 · 3 · Next

Message boards : Number crunching : S@H Enhanced v6.04


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.