Until we get more experience with volunteers running these high memory apps I think it makes sense to restrict it to a single task for now. We can change it later in light of experience.

No other projects I know of run tasks with this high memory requirements so it's not obvious how they will be received. Let's walk first before we run with this.

LHC's ATLAS tasks at 10GB are the biggest I know of. But that's 8 threads, so you don't get people trying to run huge numbers of them. Are yours going to be single threads?

Hi Kali,

The server they go to is in Hobart, NZ. I should have spotted the NZ in the task name and thought of that. Most likely when Andy gets my message he will email the data centre in Tasmania. This has happened before on a number of occasions.

Dave

Yes I am still seeing "connect(): failed" messages on all upload tries.

But I still have 4 work units running and I am no where near filling up any disks, so no problem here.

Conan

My resent task 22249228 has been sent out twice before.

Previous Task 22246540 and Task 22248943

Task 22246540 has no Stderr, it failed with a Run Time of 1 Day 5 Hours and a CPU Time of 31 Minutes. It also had an unusual amount of Peak Disk Usage of 23,961.87 MB (or 23.9 GB) way above the norm as I have seen.

Task 22248943 has the error "Process exited with code 9" other than that seemed to have run fine. This one belonged to wateroakley

I was able to run this WU to completion without error.


Another resent task I have running is Task 22249324

Previous Task 22247025 and Task 22249194

Task 22247025 on computer 1524992 it had a Run Time of 42 Minutes with a CPU Time of 20 Seconds with a Peak Disk Usage of just 404.06 MB.
This computer still has work on it but has not completed a successful OpenIFS WU all failed work units have the same long run times and short CPU times and have different error codes as well, codes 1, 5 and 148 all appear on this computer.

Task 22249194 on computer 1504810 has No Stderr, has a Run Time of 1 Day 1 Hour and CPU Time of 7 Hours.
This computer has run 9 OpenIFS work units all have failed with the long Run Time and short CPU Time.
This computer belongs to happywetter.at

So a few different reasons that some work units have failed or thrown an error.

Conan

Just downloaded a resend of a Work Unit that failed due to an error.

This Task 22245903

It failed due to running longer than 5 minutes after the work unit had finished.

The WU was run by mikey and other than the longer run time after finishing seemed to have run successfully after over 2 days run time.

The run time seems overly long on a Ryzen but did complete.

It is now running as Task 22249047 on my Ryzen computer.

Will see how it runs for me.

Conan

OpenIFS 43r3 Perturbed Surface, has been added to the application list, what does it mainly cover? Thanks
Conan

it's a modified version of the default OpenIFS in which the surface parameters, instead of the atmospheric ones, can be modified. There are two large (~3000) ensembles planned for this month where each member of the ensemble has slightly different parameters. I'll ask the scientists to write something to the forum about it.

Just for information. Hoping these do come to main site soon. One of my Open IFS tasks from testing had the following on the task page.

Peak working set size 8.77 GB
Peak swap size 9.38 GB

Oh and some of them have had final uploads in the region of 1GB too!

I expect the event log would make interesting reading.

It sure would. Too bad the user never looks at it. Is he not even curious that he has obtained no credit for years of work?

Maybe he died and no one has found out, or even turned off his machine.

@Jean-David

The install of his Fedora 36 did not include the libnsl file that is apparently needed. This results in upload failures (for some reason). He has installed this file now and the 6th and final file did upload correctly. However, since the other five monthly zip files did not go up (before he installed libnsl), boinc marked the results as errors.


@Conan

Looking at the stderr on your task webpages, the 6th zip must have been uploaded successfully, but the other 5 monthly zips weren't. So boinc marked the result as an error. Now that libnsl is installed, you shouldn't have any more errors of this type.

Yes, it's the hadam3pm2 I was concerned about. My thought is to let it go on for a while to see if its "slowness" persists. It's possible I'm wrong, or getting excited prematurely; in any case it should be clearer after a few more days.

Actually, I'm really curious to know if anyone else running the model has encountered the same issues--or not.

I'm set to get new work, but communication keeps getting deferred for one hour. In the meantime, Einstein and Seti continue working just fine.

Hi,
I found for TrendMicro, that the web-reputation service is causing the problem.
The web-reputation service doesn't allow a download from the download share you are using for the boinc program files like "hadam3p_afr_7.22_windows_intelx86.exe".
The project program files are created, but only contain the info text from web-reputation service.

Will try to post the web-reputation message from file tonight.

Australia temperatures rising faster than rest of the world: official report

Australia faced a rise in temperature of potentially more than 5 degrees celsius (41 degrees fahrenheit) by the end of the century, an increase that would outpace global warming worldwide, the country's national science agency said on Tuesday.

NB: A rise of 5 degrees C is 9 degrees F. It is not an absolute temperature. Some journalist needs to go back to school.

Also every time I visit the climate site and visit my account I have to re-loggin despite checking the "keep logged in box"... its that a cookie issue?.

I've certainly had credit totals for a host reduced when that host was merged, which anomalous behaviour was passed onto the project team (failure to carry across archived results?). However, my overall total didn't reduce.

I'm pleased so far with my new i7. I was expecting it to have four cores but was surprised to find eight.

I'd started listening to YouTube recordings of Teresa Berganza while doing various jobs on the computer but it's crunching such a lot and the eight models are using so much of my 20GB monthly bandwidth allowance that I've had to limit my listening.

http://www.youtube.com/watch?v=dWGc9IoxhAw Las tres hojas
http://www.youtube.com/watch?v=ZBEoTZvpHPg Sevillanas

It also has a usable GPU but there's not enough spare bandwidth to let the GPU have more than three or four hours' practice on Einstein each day.

I've had a reply, and I've been thinking about it for a couple of days. But I don't understand the figures, and I can't get my credits to add up either.

This is your data:

The archive.stats table says there are 36 entries for a total of 690920.619526058. The entries in the archive for the host you mention are:

    +-----------+------------------+
    |  resultid  |  total_credit  |
    +-----------+------------------+
    |  8400870  | 52254.7200679779 | 
    |  8401264  | 52254.7200679779 | 
    | 10812572 |  2081.77205586433 | 
    +------------+------------------+

The BBCode isn't working. :(
My credits are spread over several defunct parts of cpdn, and I've never kept track of what I've had, so I'm going to have to leave this for someone else to explain.

Thanks for the reply Les,

There should be a lot more results in the archive.

Working backwards with a bit of maths that may help

I have 2 hosts on this project Host 852550 and Host 1124050 .

Host 852550 has a total of 594,397 which appears to be correct.
Total viewable WUs on account total 211,409

Host 1124050 shows total viewable WUs of 184,021.

So Total archived for Host 852550 equals 594,397 - 211,409 = 382,988.

Total amount of archived points = 690,921 So 690,921 - 382,988 = 307,933.

This means that 382,988 archived points belong to Host 852550 and 307,933 archived points belong to Host 1124050.

My Total points 1,086,351 minus Archived points of 690,921 = 395,430 points.

My Total viewable WUs for both hosts is 211,409 (host 852550)+ 184,021 (host 1124050) = 395,430 points

So this bit adds up.


Now if you take my Total of 1,086,351 ans subtract the 594,397 points of Host 852550 then you are left with 491,954 points.

491,954 -
307,933 = (archived points total for Host 1124050)
184,021 (which equals the viewable amount of WUs for Host 1124050).

So I Claim that Host 1124050 should have 491,954 points (not 290,613 as displayed).

Les you show 3 WU that total 106,591.21 points from the archive

I have shown that there is 307,933 archived points for Host 1124050.

Subtract 106,591 from 307,933 you get 201,342 (rounded) which is the original amount that I asked to added back onto Host 1124050.

If you subtract displayed total of 290,613 from Actual Total of 491,954 then again you get 201,341.

So HOST 1124050 should have a TOTAL of 491,954 points.

It is Missing 201,342 (rounded) points from its total.

I don't know why you can't see the rest of the archived work units but I know the maths and I know the amount I had before any results were originally archived.

Thanks for your effort in this Les, it is much appreciated.

Hope the maths makes sense

Conan

I've had a reply, and I've been thinking about it for a couple of days. But I don't understand the figures, and I can't get my credits to add up either.

This is your data:

The archive.stats table says there are 36 entries for a total of 690920.619526058. The entries in the archive for the host you mention are:

    +-----------+------------------+
    |  resultid  |  total_credit  |
    +-----------+------------------+
    |  8400870  | 52254.7200679779 | 
    |  8401264  | 52254.7200679779 | 
    | 10812572 |  2081.77205586433 | 
    +------------+------------------+

The BBCode isn't working. :(
My credits are spread over several defunct parts of cpdn, and I've never kept track of what I've had, so I'm going to have to leave this for someone else to explain.

I did pass on a request to look at your problem, but didn't get a reply.

And with all of the problems for the last month or so, with the project virtually shut down, it's unlikely that any has been done or will be any time soon.

I noticed that about a month ago the Host Total of This Host had dropped by a large amount.

I have waited to see if the totals sort themselves out but alas they have not.

The Total should be about 464,804.47 cobblestones, however it is showing 290,612.84 cobblestones (290,613 after rounding).
Now this lower total happens to be the total of all displayed model results on my account page for this host.
All archived results after being removed from my account page took their credits with them.

My overall total appears OK and my other computer host also appears OK.

If this discrepancy can be corrected It would be much appreciated thanks.

Conan

Happy New Year to the CP team.

Have done a recalculation and of the total of 1,086,351 one host has 594,397 credits and the one with the problem has 290,613 credits.

Added together this is 885,010 credits.

So 594,397 plus 290,613 = 885,010

1,086,351 minus 885,010 = 201,341 credits missing from host total.

290,613 + 201,341 = 491,954 credits This should be my hosts total.

So Host 1 has 594,397 credits
and Host 2 should have 491,954 credits.

I only have the two hosts connected to this project so it is easy to see the missing credits on my hosts totals.

Unsure if others are having a missing credit issue.

Well, as has been posted elsewhere, there's no programmers here at the moment, and their replacements won't arrive for another few weeks.
And when they do, they have a lot of work to do that will probably be considered far more important then fiddling with credits.
So it could be six months or more before it's fixed.

I noticed that about a month ago the Host Total of This Host had dropped by a large amount.

I have waited to see if the totals sort themselves out but alas they have not.

The Total should be about 464,804.47 cobblestones, however it is showing 290,612.84 cobblestones (290,613 after rounding).
Now this lower total happens to be the total of all displayed model results on my account page for this host.
All archived results after being removed from my account page took their credits with them.

My overall total appears OK and my other computer host also appears OK.

If this discrepancy can be corrected It would be much appreciated thanks.

Conan

Happy New Year to the CP team.

That error isn't all that uncommon, and will result in part of the software marking a model name for download, and a different part NOT sending it. So your records show you as having gotten it, even though it was never sent. Most of us have a few like this.
They don't get deleted from your list.

3-4 months is typical for these models.

These models trickle early in December. (3rd or 4th, I think).
The first one will be about 0.6%, I think.

I've had several, and I don't mess around with them. If they go blue, I abort them and get another one.

I see you've now aborted that iceworld, Mican. Its whole workunit #6278040 is rather interesting. A model on a machine with Linux crashed at about the same point. The model Hagar mentioned that seems to have speeded up was on a Mac. It crashed.

Conan has a model from the same workunit running on Linux and Verstapp has one on Intel/Windows, both further behind. Both Conan and Verstapp are contactable. I'll ask them to watch their models to see what happens well into Phase 4. It will be particularly interesting to see whether Conan gets an iceworld on Linux.

G'Day mo.v,
Just an update, I took a quick look at Verstapp's work unit as it has reached phase 4 and I am still at phase 1. The last TS that he uploaded has just taken a large drop from a consistent 1.2 s/TS to 1.33 s/TS so perhaps he has also gone into an Ice world ?

I see you\'ve now aborted that iceworld, Mican. Its whole workunit #6278040 is rather interesting. A model on a machine with Linux crashed at about the same point. The model Hagar mentioned that seems to have speeded up was on a Mac. It crashed.

Conan has a model from the same workunit running on Linux and Verstapp has one on Intel/Windows, both further behind. Both Conan and Verstapp are contactable. I\'ll ask them to watch their models to see what happens well into Phase 4. It will be particularly interesting to see whether Conan gets an iceworld on Linux.

It is necessary to use the same email account. Have you changed email addresses?

The whole question of BOINC downloads and upgrades for Linux is a mess as far as I can see. Not just the stability of the BOINC version, but the whole question of how to get it going on all the different distros.

It\'s not as bad as it seems but it\'s definitely not as easy as with Windows. If there were people dedicated to the job of building BOINC install/update packages it would be great but the manpower, expertise or dedication seems to not be there.

If you do crunch a climate model again, it might be worthwhile in future holding off upgrading BOINC until you\'ve completed the last model on the machine and have nothing or nearly nothing from other projects.

Personally, I wouldn\'t dream of updating Linux or BOINC while a CPDN model is running. If it\'s absolutely necessary (and having the latest eye candy or strictly for convenience feature doesn\'t = necessary) then I make redundant backups and have a rock solid plan for rolling back the update if it causes problems. Also, I will not run CPDN parallel with other projects. If I start a model then it runs 24/7 to completion, no preempting by other projects because the sooner you get them done the fewer headaches you have. Then I put CPDN on the back burner for a few months and help other projects. Why take unnecessary risks that have zero payoff?

>snip<

The current 5.44 model can be 30% slower than the 5.15 models at the BBC. So, your previous 2 sec/timestep becomes 2.6 s/ts.

So, the numbers are about right - but always slightly worse ...

thanks, interesting point about v.5.44 against the older v.5.15 which I didn\'t know.
the calculations make sense, however we still have the two puzzles of why the models are decelerating fairly systematically and why the second model (160yrs rather than the previous 80yr one) started off at the speed the first one finished at.

Simply amazing and strange. My only guess at this point is a flaky BOINC version. I think I tried to update to 5.10.8, 5.10.10, and 5.10.20 all at different points, but the programs had issues running on FC7. The manager and the daemon had communication/launching/terminating problems. You may have discovered a bug.

BTW, are you running the 32-bit or 64-bit version of BOINC 5.10.8?

I\'d recommend doing the stability test even if the PC has been stable for a long time. There are several things which can affect your PC, such as dust on fans and heatsinks, fan bearings wearing, power supplies wearing out, and so forth.

What was the status of the 0% job in PS / top? i.e., perhaps D, or S.

Do you have \'keep in memory\' set to yes or no? If no, then turning it on may help. Sometimes if a model goes out of memory at the wrong point, it resets it.

Yes I do.

Have you tried running a stability check on the PC for 24 hours or so? (mprime is the linux version of Prime95). Note that you\'ll need to run 4 copies, one for each core.

No I have not, and seeing as the problem has just started within a week of upgrading Boinc I suspect Boinc not my PC. It has been very stable and reliable since I built it a year ago.

When you say \'one of the cores shuts down\' what does this mean? Does the job still appear in PS / Top ? What is it\'s status if it does? What does the boinc manager show for that task?

Your computers are hidden, so we can not look at the model to see if it did indeed finish.

Oh Poo :O) How about \"Granted credit 39,657.60\"? Is that what is normally granted for a finished model?

Hi Conan

At the end of June while looking at someone else\'s models we came across the first super-maxi-downloader you mention called A**. I sent him two private messages when I found he was registered on the independent forum. They\'re still sitting in my outbox which means he hasn\'t picked them up. We reported him to the programmers who were about to cut off his supply, but then Ash\'s hard drive must have filled up with initial model data (!!!) and the downloads stopped.

The programmers now have a script to automatically cut off the supply to particularly bad offenders. The script also works on members we report to them. We recently reported another megadownloader who did not reply to an email we sent. The script cut off his supply.

Unfortunately the (only) two programmers are working under pressure and don\'t want lots of these reports, only the worst cases.

When I have a moment I\'ll look at your leads to the other megamodellers to see if there\'s some way these people can be contacted. The mods have no access to members\' email addresses, but sometimes there are other legit methods. Realistically, I can only get a message to a person who

* has a website that includes an email address

* or is registered on the independent forum under a name I can recognise

* or has posted on this forum, which enables us to send an email by deleting a post and completing the form

This has been going on since at least last year.
Attempts have been made to contact the users, to no avail, so, yes, there is a list to which the computers can be added that prevents them from getting more.
If the person concerned ever becomes aware that their computer has lots of models, fixs the problem, and then wants to continue, they can post and ask for help.
And also explain what has been going on.

You\'ll notice that host 517516 stopped uploading results in mid-July, I gather that was when it was added to a blacklist.

I am so happy - after 4300 hrs I have just completed a model (believe it was an ocean).

8/14/2007 5:27:14 PM|climateprediction.net|Computation for task hadcm3ohc_182r_05605482_1 finished

It\'s been a while since I\'ve had one run the whole way through.

Wendy

Credit is a fickle thing. It can sometimes take a while if the program stops for some reason.
It\'s more like \"climate\" than \"weather\". But it\'ll show up sooner or later.
(A bit like death and taxs.)

> Have passed 200,000 cobblestones in CPDN.
Also passed 200,000 recently in Docking and Ralph.
Bit of a triple milestone.

Trying to get to 500,000 in QMC and 400,000 in Einstein.

Was hoping to finish my two current CPDN models and have a rest as that would make 3 complete models, but accidently downloaded another two models. So I will have to keep going for a bit (lots of bits) longer.

At least I now know why I can\'t upload, but why am I classed as spam?

Hold on, the rating code is something different from the anti-spam code. They don\'t work together.

The anti-spam code is checked through an external server. Your whole post is checked on words it deems spam.

The rating is something you set yourself. Check your forum preferences and see what the settings are for Filtering. Default is -25 and 5. Perhaps that you put them to -1 and 1...

Save changes to the site and you\'ll see all posts with a -1 or lower rating again. Let\'s then hope that the person or persons doing the rating get a life. ;-)

(P.S: I have mine set to -1001 and 1000... that way I see all posts in a normal way.)

Not the old FILE.
The entire BOINC FOLDER, along with all of the sub-folders.

The data needed to restart is scattered over several folders, starting in the main BOINC folder, and extending down to a sub-folder of the models folder.

The error code 1 may well be because of shutting down your computer without first exiting from BOINC, whatever operating system you use.
Or because of an older version of the graphics software.

this should be updated (i.e. Conan is founder of \"TeamAUS\"). Unfortunately there is no \"official\" BOINC way for this, so I just did it manually.

Mods can\'t help you. No tools. Don\'t know what the Board allows the Oxford staff to do -- but they are two and are buried in high-priority development/maintenance/database reorg/et al. Sorry.

To the project people,
I am submitting this post in the hope that you can help to make me the founder of the team I am in.
I have joined the team TeamAUS, I found out
the Founder of the team (Craig Zuvich), passed away suddenly on the 16/8/05 with his wife letting his computers run till the end of the year before turning them off.
TeamAUS participates in ClimatePrediction.net, Einstein@home, LHC@home and Predictor@home. I added Rosetta@home when I first joined.
I have been in and out of 3 projects in this team since October 2005 including SETI of which David at Seti, made me the founder late last year (recently I have moved this team). I am now adding LHC and ClimatePrediction to the projects that I do (I did not know the units were so big and long in CP, but I will get through them).
I would like to be made the Founder as I am the only person now in the team.
Team name is TeamAUS
Original founder is Craig Zuvich
My name is Conan
Country is Australia.
Thank you for your time.

To the Project People / Moderators
It has now been 15 days since posting, can this request be seen to please, I can\'t do anything with this team until made the Founder.
Thank you, and awaiting your reply.