Keith could you please tell me whether your Mac is overclocked or running at stock speed?

---

Cpdn news

Hi, Keith:

What happened to you (loosing 2 famous tasks in 12 hours) isn’t really that unusual. A few weeks ago I lost 2 famous models in less than 8 hours. About 1/3 of Famous WU's fail because of internal causes. I drew 2 of the less stable WU's from the server. Just bad luck. I haven’t lost one since.

---

Mo

I have just seen activity on the movement of credits, so am reassured that results are finding their destination. I normally connect between 11pm and 1am (GMT or BST) each day, but have been going on to an active network more frequently in the last week as the downloads were failing very early and needed to be replaced.

I am running at normal speed
I have never "overclocked" and have never felt inclined to do that and consequently have never discovered how to do so.
Jim has just said it is not unusual for FAMOUS tasks to fail.
(Thanks for input, Jim).
But the number of calculation errors is just about 100% at this moment.
Far above the previous average.
And all seem to have a multiple request from BOINC followed by a missing THETA statement.
If it is a test to see if the applications are working or not, I have little hope of them achieving anything from the failure rate as it is at present!!!

And I am concerned that there is virtually no action on our "number Crunching" forum.
Have we lost so many crunchers through exasperation? That concerns me, as I am seriously thinking of giving it a rest for some time to see if CPDN is going to survive.

Previously there has been an alternative to running FAMOUS, but now there is no other option, which makes it very frustrating. If the FAMOUS tasks are in an experimental stage, it would be encouraging to have some feed back to know what is being achieved from these numerous failures, otherwise I am sure I am not the only one feeling frustrated.

I see I have only 2 active tasks right now and should be getting another downloaded, but am being told I am "not reporting or requesting tasks". I just wonder if I am going to be left without my 2nd cpu having any work when the next failure occurs!!! I have had 3 failures since 10 am (i.e. In under 8 hours).

Keith

P.S. As I was told that failures can be caused by quitting OSX before quitting BOINC, I decided yesterday that I would always in future quit BOINC first, but have not yet had a system update or anything that would require me to do that as yet.

Mo

I have just (5:45pm) increased my cache size and immediately got one extra FAMOUS download, so, at least that is working!!

I will list my activity from that time for you to see what is reported there.
It shows, I think, that I am not overclocked, with a normal CPU benchmark.
It also shows that I am unsuccessful in my preferences for all but FAMOUS tasks.


Mon 13 Dec 17:29:08 2010 climateprediction.net Sending scheduler request: Requested by user.
Mon 13 Dec 17:29:08 2010 climateprediction.net Not reporting or requesting tasks
Mon 13 Dec 17:29:09 2010 climateprediction.net Scheduler request completed
Mon 13 Dec 17:44:14 2010 climateprediction.net update requested by user
Mon 13 Dec 17:44:14 2010 climateprediction.net Sending scheduler request: Requested by user.
Mon 13 Dec 17:44:14 2010 climateprediction.net Not reporting or requesting tasks
Mon 13 Dec 17:44:15 2010 climateprediction.net Scheduler request completed
Mon 13 Dec 17:45:23 2010 climateprediction.net General prefs: from climateprediction.net (last modified 04-Dec-2010 23:16:36)
Mon 13 Dec 17:45:23 2010 climateprediction.net Host location: none
Mon 13 Dec 17:45:23 2010 climateprediction.net General prefs: using your defaults
Mon 13 Dec 17:45:23 2010 Reading preferences override file
Mon 13 Dec 17:45:23 2010 Preferences:
Mon 13 Dec 17:45:23 2010 max memory usage when active: 1638.40MB
Mon 13 Dec 17:45:23 2010 max memory usage when idle: 1843.20MB
Mon 13 Dec 17:45:23 2010 max disk usage: 10.00GB
Mon 13 Dec 17:45:23 2010 don't use GPU while active
Mon 13 Dec 17:45:23 2010 suspend work if non-BOINC CPU load exceeds 75 %
Mon 13 Dec 17:45:23 2010 (to change preferences, visit the web site of an attached project, or select Preferences in the Manager)
Mon 13 Dec 17:45:26 2010 climateprediction.net Sending scheduler request: To fetch work.
Mon 13 Dec 17:45:26 2010 climateprediction.net Requesting new tasks
Mon 13 Dec 17:45:32 2010 climateprediction.net Scheduler request completed: got 1 new tasks
Mon 13 Dec 17:45:34 2010 climateprediction.net Started download of famous_v6k2_1599_200_006694264.zip
Mon 13 Dec 17:45:34 2010 climateprediction.net Started download of dump_r3u8_20a_1599.gz
Mon 13 Dec 17:45:35 2010 climateprediction.net Finished download of famous_v6k2_1599_200_006694264.zip
Mon 13 Dec 17:45:35 2010 climateprediction.net Started download of dump_r3u8_20o_1599.gz
Mon 13 Dec 17:45:37 2010 climateprediction.net Sending scheduler request: To fetch work.
Mon 13 Dec 17:45:37 2010 climateprediction.net Requesting new tasks
Mon 13 Dec 17:45:39 2010 climateprediction.net Scheduler request completed: got 0 new tasks
Mon 13 Dec 17:45:39 2010 climateprediction.net Message from server: No work sent
Mon 13 Dec 17:45:39 2010 climateprediction.net Message from server: No work is available for UK Met Office HadSM3 Slab Model
Mon 13 Dec 17:45:39 2010 climateprediction.net Message from server: No work is available for HadCM3 Coupled Model Experiment Optimised File I/O
Mon 13 Dec 17:45:39 2010 climateprediction.net Message from server: No work is available for UK Met Office FAMOUS
Mon 13 Dec 17:45:39 2010 climateprediction.net Message from server: No work is available for UK Met Office HADAM3P European Region
Mon 13 Dec 17:45:39 2010 climateprediction.net Message from server: (reached daily quota of 2 tasks)
Mon 13 Dec 17:45:43 2010 climateprediction.net Finished download of dump_r3u8_20a_1599.gz
Mon 13 Dec 17:46:05 2010 climateprediction.net Finished download of dump_r3u8_20o_1599.gz
Mon 13 Dec 17:47:58 2010 climateprediction.net Started upload of famous_v7cx_1399_200_006695303_1_1.zip
Mon 13 Dec 17:48:50 2010 climateprediction.net Finished upload of famous_v7cx_1399_200_006695303_1_1.zip


Keith

If you look at the graph above it seems there is a large spike as of today... perhaps the credits finally got counted for the past couple days? I noticed a drop in the RAC from the past couple days but I think now that it's showing up again it will compensate for that. Also, I saw uploads on some of my machines that were stalled and just started working again today. A big thanks to all the server admins who (try to) keep things running smoothly!

Also a quick question, Do sites like BoincStats pull the stats from CPDN at different intervals than CPDN posts theirs? I see variations in the data occasionally.

When CPDN updates its credits each day the credits then have to wait until the credits websites like BoincStats come along and read the figures. So there's always some delay before we see our credit updates on the stats websites.

Keith, I thought your computer wasn't overclocked so thanks for confirming that.

The reason why you can only download FAMOUS models at the moment is that the regional models are currently only available for Windows, not Mac or Linux. (Previously they were available for everyone.) I know this is a nuisance as it gives Mac and Linux users no choice at present. We hope the regional models will be available for everybody later but I don't know when.

We'll continue to post about model availability and other changes everybody needs to know about in the News thread at the top of the Number Crunching section.

---

Cpdn news