climateprediction.net home page
CPDN not using all processors

CPDN not using all processors

Questions and Answers : Windows : CPDN not using all processors
Message board moderation

To post messages, you must log in.

AuthorMessage
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1029 - Posted: 16 Aug 2004, 9:58:03 UTC
Last modified: 16 Aug 2004, 10:00:26 UTC

Today, I attached an Intel P4 HT (=2 Processors) to CPDN; it runs only on one processor (instead on two, as it should).

The box was running BOINC 3.20 / 3.22 with Seti and Predictor before.

I installed BOINC 4.x client over 3.20/3.22 without uninstalling 3.20/3.22 (as people will do, if all have to upgrade to BOINC 4.x)

After upgrading, I attached project CPDN; the client downloaded ONE WU and started crunching it.

I tried several things with preferences, but nothing changed.

I detached CPDN and reattached again; again it downloaded 1 WU and started crunching it.

Under my boxes is one more P4 with HT (the manufacturer is the same, the configuration is the same), this one is crunching fine 2 WUs the same time.

What can I do ?

This is the box, that runs only 1 WU instead of 2: <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=1181">http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=1181</a>

This is the box, that runs fine with 2 WUs: <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=77">http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=77</a>

This is the WU, that will never get returned from the box, because I detached the project: <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=11270">http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=11270</a>

Here are some logs from the message screen:

--- - 2004-08-16 11:38:31 - Starting BOINC client version 4.03 for windows_intelx86
climateprediction.net - 2004-08-16 11:38:31 - Project prefs: using your defaults
Predictor@home - 2004-08-16 11:38:31 - Project prefs: using your defaults
SETI@home - 2004-08-16 11:38:31 - Project prefs: using your defaults
climateprediction.net - 2004-08-16 11:38:31 - Host ID is 1178
Predictor@home - 2004-08-16 11:38:31 - Host ID is 20392
SETI@home - 2004-08-16 11:38:31 - Host ID is 134359
--- - 2004-08-16 11:38:31 - General prefs: from climateprediction.net (last modified 2004-08-16 11:31:01)
--- - 2004-08-16 11:38:31 - General prefs: using your defaults
climateprediction.net - 2004-08-16 11:38:31 - Resuming computation for result 0173_200026536_0 using hadsm3 version 4.02
Predictor@home - 2004-08-16 11:38:33 - Deferring communication with project for 2 days, 8 hours, 7 minutes, and 40 seconds
SETI@home - 2004-08-16 11:38:33 - Deferring communication with project for 17 minutes and 38 seconds
climateprediction.net - 2004-08-16 11:40:23 - Resetting project
climateprediction.net - 2004-08-16 11:40:25 - Detaching from project
http://climateprediction.net/ - 2004-08-16 11:41:24 - Project prefs: using your defaults
--- - 2004-08-16 11:41:25 - CPU scheduler starvation imminent; requesting more work
--- - 2004-08-16 11:41:28 - CPU scheduler starvation imminent; requesting more work
http://climateprediction.net/ - 2004-08-16 11:41:28 - Requesting 506 seconds of work
http://climateprediction.net/ - 2004-08-16 11:41:28 - Sending request to scheduler: http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi
http://climateprediction.net/ - 2004-08-16 11:41:31 - Scheduler RPC to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi succeeded
http://climateprediction.net/ - 2004-08-16 11:41:31 - General preferences have been updated
--- - 2004-08-16 11:41:31 - General prefs: from climateprediction.net (last modified 2004-08-16 11:31:01)
--- - 2004-08-16 11:41:31 - General prefs: using separate prefs for work
climateprediction.net - 2004-08-16 11:41:31 - Project prefs: no separate prefs for work; using your defaults
climateprediction.net - 2004-08-16 11:41:31 - Started download of hadsm3_4.02_windows_intelx86.exe
climateprediction.net - 2004-08-16 11:41:31 - Started download of hadsm3data_4.02_windows_intelx86.zip
climateprediction.net - 2004-08-16 11:41:44 - Finished download of hadsm3_4.02_windows_intelx86.exe
climateprediction.net - 2004-08-16 11:41:44 - Approximate throughput 86147.450578 bytes/sec
climateprediction.net - 2004-08-16 11:41:44 - Started download of hadsm3um_4.02_windows_intelx86.zip
climateprediction.net - 2004-08-16 11:42:02 - Finished download of hadsm3data_4.02_windows_intelx86.zip
climateprediction.net - 2004-08-16 11:42:02 - Approximate throughput 146955.707716 bytes/sec
climateprediction.net - 2004-08-16 11:42:02 - Started download of hadsm3se_4.02_windows_intelx86.zip
climateprediction.net - 2004-08-16 11:42:06 - Finished download of hadsm3um_4.02_windows_intelx86.zip
climateprediction.net - 2004-08-16 11:42:06 - Approximate throughput 93535.859031 bytes/sec
climateprediction.net - 2004-08-16 11:42:06 - Started download of 017v_200026564.zip
climateprediction.net - 2004-08-16 11:42:09 - Finished download of hadsm3se_4.02_windows_intelx86.zip
climateprediction.net - 2004-08-16 11:42:09 - Approximate throughput 126783.546099 bytes/sec
climateprediction.net - 2004-08-16 11:42:09 - Finished download of 017v_200026564.zip
climateprediction.net - 2004-08-16 11:42:09 - Approximate throughput 3440.326531 bytes/sec
climateprediction.net - 2004-08-16 11:42:10 - Starting computation for result 017v_200026564_0 using hadsm3 version 4.02
Predictor@home - 2004-08-16 11:48:45 - Started upload of t0224C_1_3539_0_0
Predictor@home - 2004-08-16 11:48:48 - Temporarily failed upload of t0224C_1_3539_0_0
Predictor@home - 2004-08-16 11:48:48 - Backing off 1 hours, 54 minutes, and 36 seconds on transfer of file t0224C_1_3539_0_0

<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1029 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1034 - Posted: 16 Aug 2004, 10:44:41 UTC

It looks like the client tries to get more work from Seti and Predictor, but doesn't ask CPDN ...

My resource share at this time says:

66,66% Predictor
33,33% CPDN
00,01% Seti

Could this be the reason ? But, if so, is it good to let one processor idle, when other projects can't distribute WUs ?

Here from the message-screen:

--- - 2004-08-16 11:56:12 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 11:56:12 - Requesting 86 seconds of work
SETI@home - 2004-08-16 11:56:12 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 11:56:15 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 11:56:15 - No schedulers responded
SETI@home - 2004-08-16 11:56:15 - Deferring communication with project for 1 minutes and 0 seconds
--- - 2004-08-16 11:57:18 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 11:57:18 - Requesting 86 seconds of work
SETI@home - 2004-08-16 11:57:18 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 11:57:21 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 11:57:21 - No schedulers responded
SETI@home - 2004-08-16 11:57:21 - Deferring communication with project for 1 minutes and 0 seconds
--- - 2004-08-16 11:58:22 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 11:58:22 - Requesting 86 seconds of work
SETI@home - 2004-08-16 11:58:22 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 11:58:25 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 11:58:25 - No schedulers responded
SETI@home - 2004-08-16 11:58:25 - Deferring communication with project for 1 minutes and 0 seconds
--- - 2004-08-16 11:59:26 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 11:59:26 - Requesting 86 seconds of work
SETI@home - 2004-08-16 11:59:26 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 11:59:29 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 11:59:29 - No schedulers responded
SETI@home - 2004-08-16 11:59:29 - Deferring communication with project for 1 minutes and 0 seconds
--- - 2004-08-16 12:00:30 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 12:00:30 - Requesting 86 seconds of work
SETI@home - 2004-08-16 12:00:30 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 12:00:33 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 12:00:33 - No schedulers responded
SETI@home - 2004-08-16 12:00:33 - Deferring communication with project for 1 minutes and 0 seconds
--- - 2004-08-16 12:01:34 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 12:01:34 - Requesting 86 seconds of work
SETI@home - 2004-08-16 12:01:35 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 12:01:38 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 12:01:38 - No schedulers responded
SETI@home - 2004-08-16 12:01:38 - Deferring communication with project for 2 minutes and 12 seconds
--- - 2004-08-16 12:03:51 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 12:03:51 - Requesting 86 seconds of work
SETI@home - 2004-08-16 12:03:51 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 12:03:54 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 12:03:54 - No schedulers responded
SETI@home - 2004-08-16 12:03:54 - Deferring communication with project for 6 minutes and 7 seconds
--- - 2004-08-16 12:10:02 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 12:10:02 - Requesting 86 seconds of work
SETI@home - 2004-08-16 12:10:02 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 12:10:05 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 12:10:05 - No schedulers responded
SETI@home - 2004-08-16 12:10:05 - Deferring communication with project for 2 minutes and 57 seconds
--- - 2004-08-16 12:13:03 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 12:13:03 - Requesting 86 seconds of work
SETI@home - 2004-08-16 12:13:03 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 12:13:06 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 12:13:06 - No schedulers responded
SETI@home - 2004-08-16 12:13:06 - Deferring communication with project for 15 minutes and 20 seconds
Predictor@home - 2004-08-16 12:21:53 - Started upload of t0241C_1_2826_1_0
Predictor@home - 2004-08-16 12:21:56 - Temporarily failed upload of t0241C_1_2826_1_0
Predictor@home - 2004-08-16 12:21:56 - Backing off 1 hours, 32 minutes, and 41 seconds on transfer of file t0241C_1_2826_1_0
--- - 2004-08-16 12:28:27 - CPU scheduler starvation imminent; requesting more work
SETI@home - 2004-08-16 12:28:27 - Requesting 86 seconds of work
SETI@home - 2004-08-16 12:28:27 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-08-16 12:28:30 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-08-16 12:28:30 - No schedulers responded
SETI@home - 2004-08-16 12:28:30 - Deferring communication with project for 44 minutes and 24 seconds
Predictor@home - 2004-08-16 12:38:33 - Deferring communication with project for 2 days, 7 hours, 7 minutes, and 40 seconds

<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1034 · Report as offensive     Reply Quote
Profile old_user83

Send message
Joined: 5 Aug 04
Posts: 30
Credit: 39,745
RAC: 0
Message 1037 - Posted: 16 Aug 2004, 11:49:12 UTC - in response to Message 1034.  

Hi Yeti,

Hmmmm, an interesting one. When I look at your “Computers belonging to Yeti” display I find the following:

<img src="http://cpdn.tuxie.org/mikeymoss/Yeti_Computers_BOINC_CPDN_View_20040816.jpg">

It looks like there are some duplicate entries in here to me? Maybe you can confirm?

You can certainly merge computers with other hosts (computers) or delete hosts. Maybe this is what you need to do? I think you need to reconcile what WU’s are associated with what hosts and allocate accordingly. Ideally for a 2 CPU (HT) box then allocate two WU’s, unless you want to time-share. It seems to me you know what you’re doing and maybe detach any surplus WU’s from CPDN/BOINC. What would be interesting is to see your projects and work screens from the BOINC GUI. Ultimately this should tell you if you’ve got too many projects/WU’s allocated.

AFAIK, if you’re set up to use two processors (as per HT/Dual) and so that is a setting of “2” in the “On multiprocessors, use at most” setting of Preferences -&gt; General @ http://climateapps2.oucs.ox.ac.uk/cpdnboinc/home.php (your user) then you should be OK to go. It seems to me you already have this setting enabled.

Additionally, try looking at these two threads and you may find some information of interest, where there seems to be some tips for running CPDN Classic and BOINC for your type of machine:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=10
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=58
<img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=11">
<i> UK4CP @ www.uk4cp.co.uk (United Kingdom Group) Celeron 2.6GHz XP Pro SP1 768MB RAM<i>
ID: 1037 · Report as offensive     Reply Quote
Ingleside

Send message
Joined: 5 Aug 04
Posts: 108
Credit: 18,243,376
RAC: 35,311
Message 1038 - Posted: 16 Aug 2004, 11:49:17 UTC

Increase the resource-share for CPDN and you will get work. Afterwards, you can decrease the share again.
ID: 1038 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1041 - Posted: 16 Aug 2004, 12:38:27 UTC - in response to Message 1037.  

&gt; Hi Yeti,
&gt;
&gt; It looks like there are some duplicate entries in here to me? Maybe you can
&gt; confirm?
&gt;

Hi Mikey,

I'm sorry, but your guess was wrong. All machines, you have seen in my profil, really exist and are all really crunching CPDN ... (Together with my laptop and the pc from my son, I have a farm of 14 boxes)

The settings in my profiles are set up correctly, because Predictor and Seti crunched 2 WUs simultaneous ...

I guess, Ingleside will be right, that I have to raise the resource-share, because I can see, that the client is asking Seti for more work ...

I'm a little bit disappointed, because my hope was, if Seti and Predictor can't send WUs, the client should order one from the next project, = CPDN


<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1041 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1043 - Posted: 16 Aug 2004, 12:50:23 UTC - in response to Message 1038.  

&gt; Increase the resource-share for CPDN and you will get work. Afterwards, you
&gt; can decrease the share again.
&gt;

This was right ! First I raised the resource-share from CPDN to same value as Predictor, nothing happened. Next, I raised CPDN higher than Predictor and I got the second WU for this machine from CPDN

I don't know, wheather this is what boinc-developer wanted, but from User-Side I expected that the client should download the second WU from CPDN, when Seti and Predictor can't send WUs, without changing the resource-share.

Thanks to Ingleside

<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1043 · Report as offensive     Reply Quote
Jord
Avatar

Send message
Joined: 5 Aug 04
Posts: 250
Credit: 93,274
RAC: 0
Message 1051 - Posted: 16 Aug 2004, 14:20:20 UTC
Last modified: 16 Aug 2004, 14:30:59 UTC

But why would you want to have any of the other projects send you work units? By the time your CPDN WU has finished, the deadline for the other projects is long past.

Then there's the effect that neither Predictor, nor Seti nor Pirates is using the 4.xx client yet. So they'll contact their various servers only to be told that the BOINC client in use is too high for their requested version. Then they will go into deferral mode until at one point in time those projects change to the new client. That can be a long wait. Or you must like deferrals of several weeks, hours, minutes and seconds. ;)

Oh, if you think you can run one project on one CPU while another uses the other CPU, then be advised that BOINC doesn't work that way. (Yet)
So it is not possible to use half your CPU or one physical CPU on CPDN while the other half or the other physical one crunches PAH. Maybe that a future CPU can do this, but until that time, the project you are crunching will just download two WUs for each CPU (half) to crunch one at the same time.
--------------------
Jordâ„¢
<img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=2&amp;trans=off">
ID: 1051 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1052 - Posted: 16 Aug 2004, 14:42:10 UTC
Last modified: 16 Aug 2004, 14:45:50 UTC

Hi Jord,

I know, that the other projects can't deliver work-units at the moment, because I'm running BOINC 4.x and they still use BOINC 3.x. But, as to read from startpage of CPDN:

Unfortunately this is not yet compatible with SETI and Predictor@Home, <b><i>but soon BOINC v.4.02+ should be ready for public use, project sharing, etc.</i></b>

The client 4.x is designed to use project-sharing in a different way than the 3.x version. CPDN makes it necessary, that the resource-share is switched from WU after WU thinking in a timeline-thinking and the client will have to switch crunching from project to project based on Hours or similar. So, even if all processors have a CPDN-WU to crunch, the client 4.x will be able to participate in Seti / Predictor and others, respekting all their deadlines.

This will bring a great effort to the whole boinc-project, because all the time, the other projects can't support their users with enough WUs, the client can continue with CPDN. So, there will be no idle clients, and all projects get the best support they can get.

Hopefully, it takes only several days until Seti and Predictor upgrade to BOINC 4.x

I agree to you, that, as long as Seti and Predictor are on 3.x, it is not the optimum, having the client 4.x asking the 3.x-Server for work. Most of the time, I disable network-access on the 4.x clients ...

But if they are on 4.x, I will crunch Predictor and Seti same time as crunching CPDN. That is, what the 4.x client is designed for.

EDIT: What we are doing here is beta-testing. I'm trying to use and understand the client 4.x as it is supposed to work to find out, wheather it works as it should ...

<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1052 · Report as offensive     Reply Quote
Jord
Avatar

Send message
Joined: 5 Aug 04
Posts: 250
Credit: 93,274
RAC: 0
Message 1056 - Posted: 16 Aug 2004, 15:32:20 UTC - in response to Message 1052.  

&gt; EDIT: What we are doing here is beta-testing. I'm trying to use and understand
&gt; the client 4.x as it is supposed to work to find out, wheather it works as it
&gt; should ...
&gt;
I knew that, the beta testing. ;) It's what I signed up for, weirdly enough. :D

Anyway, as long as the other projects aren't switched over to the 4.xx client/server side, you can try all you want to see if it works as it should, but will never get any results of it. ;)

As far as the front page goes... it's being rebuild as we speak, so I can't find anything anymore there. ;)
--------------------
Jordâ„¢
<img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=2&amp;trans=off">
ID: 1056 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1059 - Posted: 16 Aug 2004, 15:45:06 UTC - in response to Message 1056.  

&gt; Anyway, as long as the other projects aren't switched over to the 4.xx
&gt; client/server side, you can try all you want to see if it works as it should,
&gt; but will never get any results of it. ;)

I don't think so: The problem, not getting the second WU, is something interesting. I don't know, wheather it is as it should, but I hope, someone from develop-team will read this ...

And there is something more interesting: The projects Predictor and Seti have been resetted; the work I have done the last day will never be credited. I expected this and its okay. But, the client is still trying to bring the old ready-crunched results back to the project-server. So, the reset didn't clear all

&gt; As far as the front page goes... it's being rebuild as we speak, so I can't
&gt; find anything anymore there. ;)

Yep, noticed this also


<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1059 · Report as offensive     Reply Quote
Ingleside

Send message
Joined: 5 Aug 04
Posts: 108
Credit: 18,243,376
RAC: 35,311
Message 1074 - Posted: 16 Aug 2004, 19:51:24 UTC - in response to Message 1051.  

The client will wait 24 hours if wrong major version, that is then you get a successful connection, and this shouldn't be a huge problem compared to everyone sitting on "retry". ;)

Has seen that v4 will successfully pause CPDN, make more slots &amp; start crunching other work, based on resource-share.
Also, the scheduling-server gets info on the resource-share so if you're not using insanely high cache-settings you will normally finish the work before deadline.
Ok, if you adds more projects later on or changes the resource-share you can break the deadline, but normally this will not happen.

As for the actual problem here, maybe multi-cpu still isn't correctly taken care of by the client before finding out what project to ask...
ID: 1074 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 5 Aug 04
Posts: 170
Credit: 9,705,926
RAC: 13,871
Message 1075 - Posted: 16 Aug 2004, 19:58:22 UTC - in response to Message 1074.  

&gt; Has seen that v4 will successfully pause CPDN, make more slots &amp; start
&gt; crunching other work, based on resource-share.
&gt; Also, the scheduling-server gets info on the resource-share so if you're not
&gt; using insanely high cache-settings you will normally finish the work before
&gt; deadline.
&gt; Ok, if you adds more projects later on or changes the resource-share you can
&gt; break the deadline, but normally this will not happen.
&gt;
&gt; As for the actual problem here, maybe multi-cpu still isn't correctly taken
&gt; care of by the client before finding out what project to ask...
&gt;

Reading this, it makes me really happy, that there is someone out there, who has seen all the good things i've heared about :-))

So, we only have to wait until Predictor and Seti go up to 4.x :-)

<a href="http://www.boinc.dk/index.php?page=user_statistics&amp;project=cpdn&amp;userid=34"><img src="http://34.cpdn.sig.boinc.dk?142"></a><img src="http://75.pah.sig.boinc.dk?142"><img src="http://4238.sah.sig.boinc.dk?142">
ID: 1075 · Report as offensive     Reply Quote
Jord
Avatar

Send message
Joined: 5 Aug 04
Posts: 250
Credit: 93,274
RAC: 0
Message 1082 - Posted: 17 Aug 2004, 0:40:17 UTC

Then don't read it that way, for I have the gut feeling it won't happen or be available for the next 4 to 6 months yet..

Who said your interpretation of how B4.xx would do what you wanted it to do? I take it you're not an Alhpa tester?

So then how do you know for sure that all the options you want are included in the 4.0x client? Remember that the 4.02 client is only a beta test client. :)
--------------------
Jordâ„¢
<img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=2&amp;trans=off">
ID: 1082 · Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 5 Aug 04
Posts: 172
Credit: 4,023,611
RAC: 0
Message 1083 - Posted: 17 Aug 2004, 1:23:56 UTC - in response to Message 1082.  

&gt; Then don't read it that way, for I have the gut feeling it won't happen or be
&gt; available for the next 4 to 6 months yet..
&gt;
&gt; Who said your interpretation of how B4.xx would do what you wanted it to do? I
&gt; take it you're not an Alhpa tester?
&gt;
&gt; So then how do you know for sure that all the options you want are included in
&gt; the 4.0x client? Remember that the 4.02 client is only a beta test client. :)
&gt; --------------------
&gt; Jordâ„¢
The Alpha project has not been helping much lately, as it has been down.


jm7
ID: 1083 · Report as offensive     Reply Quote

Questions and Answers : Windows : CPDN not using all processors

©2024 climateprediction.net