home page
Posts by DJStarfox

Posts by DJStarfox

1) Message boards : Number crunching : Big models (Message 63463)
Posted 2 Feb 2021 by DJStarfox
Ya know... rather than making each model so big (multiple GB per task), could the programmers simply have each task share more files in common? That way each model takes less space, if that makes sense.
2) Message boards : Number crunching : Please fix the deadlines! (Message 63095)
Posted 4 Dec 2020 by DJStarfox
If you have 1000 credit on project 1 and 1,000,000 credit on project 2, Bonic will try to have project 1 "catch up" to project 2.

If what you say is true, then that design is faulty. The resource share should try to equalize the recent average credit (RAC), not total credit.
3) Message boards : Number crunching : New work Discussion (Message 63018)
Posted 25 Nov 2020 by DJStarfox
I got five (5) of those HadAM4h tasks, but each one is consuming ~4GB of disk space, which far exceeds the 10G max I have in BOINC settings. Seems like a bug in the task scheduler...

I had to abort a few to avoid running out of disk space on the /var partition. :(
4) Message boards : Number crunching : Preferences - project options missing (Message 60943)
Posted 19 Sep 2019 by DJStarfox
I too am surprised the application selection feature is completely gone now.

Various old models would take up 10's of gigabytes. At various points, I would have to periodically delete old models in order to download any new units. By selecting only a few models, this was less of a problem.

Also, in the past, the researchers were able to use different platforms as a way to validate a model. Since floating point calculations, etc. varied slightly, it helped them tune the models with a bigger variety of data points. If such an emergent result is no longer useful, then I would highly recommend they post the minimum requirements per platform. (For example, these could go on the CPDN homepage next to the how to join section.)
5) Message boards : Number crunching : What Happened ??? (Message 58153)
Posted 25 Apr 2018 by DJStarfox
The entire website was down for an extended time. I visited several times but did not see any notice about downtime.
6) Message boards : Number crunching : What Happened ??? (Message 58148)
Posted 24 Apr 2018 by DJStarfox
When this site shuts down, people switch to the BOINC site, in particular the Projects section, top post, which is: News on Project Outages, where a message, usually from Andy, is posted.

In this case, a new thread was also created: CPDN project going offline this afternoon, which is full of posts from people, including one that I put there near the end of the thread, which I felt explains things well enough.

First of all, I (and likely many others) did not know there was a specific forum on some other site, somewhere, where a cryptic message about some maintenance was posted.... Furthermore, the site has been down for over a month! That sounds much worse than some routine maintenance, so I consider that post deceptive.

I'm glad the site is back, and I hope such a long outage does not happen again. In the future, a post on the HOME PAGE of would be much more appropriate and visible to the community.
7) Message boards : Number crunching : Volunteer to work with the BOINC Release Manger on Linux Instructions? (Message 57862)
Posted 26 Feb 2018 by DJStarfox
I'm most familiar with RHEL/Centos 6 and 7 and have been running BOINC on Red Hat based distributions for 8+ years. I'm sure i could dedicate an hour or two to write some documentaion for BOINC. Please have Richard reach out to me if he needs anything.
8) Message boards : Number crunching : Site Back Up (Message 57824)
Posted 19 Feb 2018 by DJStarfox
Anyone know why the site was offline for so long?
9) Message boards : Number crunching : New work Discussion (Message 57684)
Posted 21 Jan 2018 by DJStarfox
Speaking of new work, I just got a single hadcm3s model after many months of nothing. Seems very random. Now, the server's queue is 27 but I have no idea which model these are.
10) Questions and Answers : Unix/Linux : Shutting down for re-boot. (Message 57337)
Posted 7 Nov 2017 by DJStarfox
Dave, I've had better luck with the following global compute preference checked:
"Leave non-GPU tasks in memory while suspended" = yes

If that is disabled on your account, try enabling it and see if that improves things.
11) Message boards : Number crunching : Credit Status (Message 57239)
Posted 28 Oct 2017 by DJStarfox
This thread is too long, and after reading it, I still don't know why so many people lost all their credits in the last day or so. I lost 1.1 million credits today.

What the hell is going on guys? This needs to get fixed ASAP.

And who programed BOINC to be this volatile that a rogue process or sysadmin can just wipe out credits granted to every user? This has got to change if BOINC is to continue long-term.
12) Message boards : Number crunching : Disk Quota exeded. (Message 56376)
Posted 12 Jun 2017 by DJStarfox
I had about 1 GB upload this morning. :)
13) Message boards : Number crunching : Disk Quota exeded. (Message 56367)
Posted 11 Jun 2017 by DJStarfox
Got a dozen or so stuck file transfers. This hurts my other projects, but I've suspended all network activity until this is resolved.

Even if they get the file move started on Monday, I expect the transfer to take many hours. So, don't expect much free space until mid-day Tuesday.

BTW, why don't the programmers know how much data each model needs to upload? They should tell the sysadmin this information #models x upload size x #uploads. That way, the sysadmin can check if they'll run out of space before the tasks are finished.
14) Message boards : Number crunching : Model uploaded/finished/reported, but still in progress on the web (Message 55445)
Posted 7 Jan 2017 by DJStarfox
I've got a model that I reported as abort, long ago. But the website says still in progress:

This seems to be a systemic problem with CPDN.
15) Message boards : Number crunching : Status Page Has Lost Work Units (Message 55311)
Posted 6 Dec 2016 by DJStarfox
The status page is showing 322 tasks ready to send, but the entire sum of WU types is zero. Also, when BOINC contacts the project, it says there are no tasks to send. So, status page must be wrong, as no work is available.

16) Message boards : Number crunching : Prevent BOINC requesting GPU work ? (Message 54812)
Posted 21 Sep 2016 by DJStarfox
exclude_gpu doesn't work on 6.10... and the Wiki does not say what version you must have for it to work. :(

You use an <exclude_gpu> in your cc_config file. It allows you to prevent a gpu from being used by a particular project. Its a client side things, doesn't require any change on the server.

My cc_config file:


You can see I have excluded my one and only GPU from being used by both Climate Prediction and Einstein. This particular machine uses the GPU for GPUgrid work. You can have multiple <exclude_gpu> statements.

If you don't already have a cc_config.xml file you'll need to create it in the BOINC data directory. Use a text editor such as Notepad and make sure its an .xml file (notepad likes to put .txt on the end of the file names). After doing this either re-read the config file (on the menu bar Advanced -> Read config file) or shut it down and start it up again.

Details about the config file can be found here

17) Questions and Answers : Unix/Linux : Safe shutdown (Message 54687)
Posted 22 Aug 2016 by DJStarfox
I would recommend shutting down BOINC remotely, then shutting down the system. BOINC will close all the task cleanly as possible.

For example:
/sbin/service boinc-client stop
/sbin/shutdown -h 1 &
18) Message boards : Number crunching : wah2_eu25 work units are BIG (Message 54302)
Posted 14 Jun 2016 by DJStarfox
One behavior I've noticed.... If each WU requires 4GB of disk space and I have allocated 5GB to BOINC, then it will download 2 WU... resulting in out of disk space when trying to get more work for additional cores...

And seriously... 4GB per WU? Are you guys kidding me?

BOINC 6.10.45
19) Message boards : Number crunching : ANOTHER UPLOAD PROBLEM (Message 54020)
Posted 27 Apr 2016 by DJStarfox
A few uploads went through on Monday, but the rest of the transfers are stuck w/ a connection time-out.
20) Message boards : Number crunching : Total Credit (Message 53889)
Posted 4 Apr 2016 by DJStarfox
OK, i guess i misunderstood. I'll check my account tomorrow and see if the situation as changed.

Next 20