1) Forums : Technical Support : No Validation (Message 6436)
Posted 8 Jul 2008 by Profile Webmaster Yoda
Post:
571 results pending here. I can\'t remember a time I\'ve had so many waiting on this project.

At the rate we\'re going, I would not be surprised if we get a database meltdown in the near future... Maybe it\'s time to follow brandontroy\'s lead and set Cosmology@home to no new work.

2) Forums : Technical Support : Work units not being reissued? (Message 6199)
Posted 17 May 2008 by Profile Webmaster Yoda
Post:
Perhaps it\'s also why so I have so many work units pending (over 350 results to date and increasing). Wait and see.
3) Forums : Technical Support : No stats update ? (Message 6162)
Posted 13 May 2008 by Profile Webmaster Yoda
Post:
3) Getting work, have your CPU spend time crunching, reporting but getting no stats exporting so that your RAC drops, no word from project admins about it...

It\'s probably just something that\'s been overlooked but for stats watchers it\'s a pain to see their overall daily output drop...
4) Forums : Technical Support : No stats update ? (Message 6131)
Posted 12 May 2008 by Profile Webmaster Yoda
Post:
Still no XML Stats Exports?
5) Forums : Technical Support : I\'m Not getting any work! (Message 6013)
Posted 26 Apr 2008 by Profile Webmaster Yoda
Post:
I am running on a Mac mini


According to http://www.cosmologyathome.org/apps.php there is no Cosmology Application for Mac yet...

If that\'s the case, it\'s no surprise that Phil is not getting work on the Mac

But my computers (64bit Linux on Intel Core 2 Duo/Quad and AMD Athlon 64) are often told there\'s no work either.
6) Forums : Technical Support : Upload/Download Server Problems (Message 4637)
Posted 24 Jan 2008 by Profile Webmaster Yoda
Post:
I too decided to just detach and re-attach on my 7 computers. May have lost about 24 hours of fully and partially crunched work units but I can live with that.

I did go through and update client_state.xml etc on a couple of computers and managed to upload and report, but can\'t be bothered doing it on all of them.
7) Forums : Technical Support : Upload/Download Server Problems (Message 4595)
Posted 24 Jan 2008 by Profile Webmaster Yoda
Post:
Uploads are working fine for me now, but reporting has some issues. Here is a transcript of one attempt to update on one computer:

Project	Date	Message
Cosmology@Home	24/01/2008 12:45:41 PM	Sending scheduler request: To report completed tasks
Cosmology@Home	24/01/2008 12:45:41 PM	Reporting 1 tasks
Cosmology@Home	24/01/2008 12:45:46 PM	Scheduler RPC succeeded [server version 601]
Cosmology@Home	24/01/2008 12:45:46 PM	Message from server: [b]Incomplete request received[/b].
Cosmology@Home	24/01/2008 12:45:46 PM	Deferring communication for 6 sec
Cosmology@Home	24/01/2008 12:45:46 PM	Reason: requested by project

8) Forums : Technical Support : Upload/Download Server Problems (Message 4582)
Posted 24 Jan 2008 by Profile Webmaster Yoda
Post:
It worked for me (on one computer - several more to go) as follows:

1. Stop BOINC
2. Open client_state.xml with a text editor
3. Search and replace \"cosmos.astro.uiuc.edu\" with \"www.cosmologyathome.org\"
4. Save and close the file
5. Restart BOINC

This is a tedious process though, especially for people who are unfamiliar with XML or don\'t have access (or time) to do this on all their computers.

Is there perhaps a possibility of getting a temporary redirect from the (incorrect) url so it goes to the right one?
9) Forums : Technical Support : Pending Because Work Unit Remains "Unsent" (Message 2249)
Posted 23 Aug 2007 by Profile Webmaster Yoda
Post:
I have quite a few pending on my Pentium E2160 running 64 bit Linux as well. In case that's also a HR issue - these Pentiums are Core 2 Duos with less L2 Cache and should be able to be paired with other Core 2 processors. Some of these results have been sitting there for 4 weeks already.

But perhaps there's just not enough active Core 2 machines running 64 bit Linux.

See Host 1947
10) Forums : Technical Support : Host flying solo (Message 1174)
Posted 20 Jul 2007 by Profile Webmaster Yoda
Post:
My AMD turion X64 laptop is in the same boat, in that every WU is going in to pending and not being sent out to another cruncher.

Computer No is 710.


Perhaps it could be put in the same HR class as Athlon 64 / X2 and even Opteron? They're the same family...
11) Forums : Technical Support : WU limit discrepancy (Message 1069)
Posted 16 Jul 2007 by Profile Webmaster Yoda
Post:
It seems that the general consensus is that a higher max-wu-in-progress limit would be appreciated. I'm OK with upping the limit within reason.

The main reason why I want a max-wu-in-progress limit is that I don't want several hundred WUs tied up on one machine. It just doesn't make any sense, especially since the initial replication is 2 now instead of 4.

Is a limit of 150 or so OK?


It would be an improvement, though I recognise the need to not have work tied up on one computer for days. Maybe that could be covered by a relatively short deadline though.

I'd like to see it tied to number of CPUs available on a host, perhaps with an upper limit based on the host's duration correction factor.

Reasoning: 50 work units on a slow single core machine might take 2 days to complete, while 50 on a fast quad core (let alone dual quad) might not even last an hour.

Example: My own E4400 (at 3 GHz) does 20 work units an hour (on average) and that's only a dual core. It ran out of work within a few hours when the server developed problems. A dual quad at that speed might go through 80 in an hour.

But of course, the project's needs do come first!
12) Forums : Technical Support : can't attach shared memory (Message 1039)
Posted 16 Jul 2007 by Profile Webmaster Yoda
Post:
I've reported it in that thread now. Wait and see...
13) Forums : Technical Support : URGENT Problems Thread (Message 1037)
Posted 16 Jul 2007 by Profile Webmaster Yoda
Post:
Unable to report results or get new work, as reported in the thread "can't attach shared memory".

Typical message:

7/16/2007 12:37:07 PM Cosmology@Home Requesting 164127 seconds of new work, and reporting 30 completed tasks
7/16/2007 12:37:12 PM Cosmology@Home Scheduler RPC succeeded
7/16/2007 12:37:12 PM Cosmology@Home Message from server: Server error: can't attach shared memory
7/16/2007 12:37:12 PM Cosmology@Home Deferring communication for 1 hr 0 min 0 sec
7/16/2007 12:37:12 PM Cosmology@Home Reason: project is down


(Times are GMT+8)
14) Forums : Technical Support : can't attach shared memory (Message 1035)
Posted 16 Jul 2007 by Profile Webmaster Yoda
Post:
Anyone else getting this error


Yep. You beat me to it. Happening on all my hosts too - I have approx. 100 results to report, 200 still in cache (which might last a couple of hours, tops)

15) Forums : Technical Support : Download Errors ... !!! (Message 1006)
Posted 14 Jul 2007 by Profile Webmaster Yoda
Post:
http://cosmos.astro.uiuc.edu/cosmohome/workunit.php?wuid=166448 ... Another 1 that was sent to 2 of my PC's & erred out when Downloading on both of them ... :)

I can't for the life of me figure out why you would be getting as many download errors as you do.

The only thing I can think of is that you should change client versions.


I've had some too lately and am not using the same BOINC client as PoorBoy. I checked one of them and found that three participants had the download error. Nobody has been able to get this result working yet.

http://cosmos.astro.uiuc.edu/cosmohome/workunit.php?wuid=178964

WU download error: couldn't get input files:
<file_xfer_error>
<file_name>params_071407_042850_0.ini</file_name>
<error_code>-119</error_code>
<error_message>MD5 check failed</error_message>
</file_xfer_error>

Given it happened to all three participants, I'd be inclined to think the problem lies somewhere at your end of the connection?

16) Forums : Technical Support : Comments on Fixed Credit System (Message 996)
Posted 14 Jul 2007 by Profile Webmaster Yoda
Post:
To quote Ray Murray:
There's a fine balancing act of not short-changing your crunchers of credit where it is due and over inflating credit to where it becomes incomparable with other projects.


Indeed. It's just a matter of which projects you compare with. Do you want to be at the low end of the scale, thus attracting mostly people who are in it for the science?

Granting less than 3 credits per result may be fair when compared to BOINC benchmarks but it will be at the low end of projects that do not use those benchmarks for credit calculation. My computers get more than they would claim (with the standard benchmarks formula) at several other projects.

I'd be happier with 4 to 4.5, which is IMHO not over-inflated.
17) Forums : Wish list : Thoughts on longer, check-pointed WUs (Message 905)
Posted 10 Jul 2007 by Profile Webmaster Yoda
Post:
Do you mean the project switch interval? Default on the write to disk is 60 seconds.

Or am I misreading/missing something?


Ah, I got confused... Thought the default was 10 minutes (600 seconds).

Still, checkpointing every 10 or so minutes wouldn't hurt. Computers crash, get switched off, rebooted etc.
18) Forums : Wish list : Thoughts on longer, check-pointed WUs (Message 901)
Posted 10 Jul 2007 by Profile Webmaster Yoda
Post:
IMHO Checkpointing should be implemented if result run length goes beyond the default "save to disk at most every" setting on an average Pentium 4 or equivalent.

As for run length, I don't particularly like work units that run for longer than a couple of hours. Things can go wrong and work units crash. Losing an hour's work occasionally is not a big issue but I hate it when my machines have been crunching for many hours, only to have a work unit crash and burn. Those results are (in most cases) no use to the project either.

Depending on how your application works, running longer might also mean higher memory requirements?
19) Forums : Technical Support : Linux/Windows speed difference? (Message 893)
Posted 9 Jul 2007 by Profile Webmaster Yoda
Post:
Now, it's certainly possible that 1.19 is optimized for AMD CPUs, since I compiled it on an AMD machine. I think ifort may have set optimizations for AMD as the default when I installed it, so I'll have to look through the options to see what the deal is.


I have a dual-booting Core 2 Duo (at 2.66GHz). In Windows it takes up to 20 minutes per result. In 64 bit Linux it's rarely more than 8 minutes.

However, my Athlon 64 3700+ (at 2.6GHz) running Windows XP takes close to 20 minutes too. My guess is that it will still be faster under Linux but I haven't got a spare HDD to try it (might run it as a DRBL node tomorrow to see if there is a big difference)
20) Forums : Wish list : Linux Claims Low Credit (Message 574)
Posted 29 Jun 2007 by Profile Webmaster Yoda
Post:
I have a strong preference for projects that have fixed credits, or some method of granting credit independently of the BOINC benchmarks.

Apart from blatant cheating and mis-configured hosts, there is of course the disparity between Linux and Windows when relying on BOINC benchmarks. Fixed credit / Flops counting sorts that out too.

We end up with a level playing field, fair competition. The only way to get ahead in the competition (which many participants thrive on) is to do more work or do it more efficiently.


Next 20