Running BOINC 7.6.22.
I'm not sure what causes these- low swap space/hard drive?, etc.?, other trouble? but after reaching first "trickle", my two hadam3p_anz(etc) processes on my (2007 desktop iMac, running OS 10.11.3) will occasionally briefly work recently, but more often will tell me that they're postponing for 10 minutes (600 seconds)- and then after 10 minutes will tell me that they're postponing for 600 seconds again.

It didn't take me long to give in to the temptation of downloading some Collatz Conjecture BOINC processes just so as to give my computer & BOINC something to -do- in the meanwhile- maybe that's another part of the problem too, sorry...- but whenever I suspend those, or most of them, and return control to Climate Prediction, it's almost immediate return to "postponed"... Any ideas why? At least, anyone else seeing this lately (I see that a search of these forums turns up nothing)? Thanks.

Eric Schissel

I'm not running a Mac, and only have one PC downloading work, with a much older boinc version, but there have been problems with downloading work from the server over the last several days. Connections timing out, download/upload errors, etc.

I've had a fair number of those too, I think- but this one happens when the task is already completely downloaded (and even finished enough to have sent its first progress-"trickle")... which is the odd bit to me...
(... I think I probably sound ungrateful to the admins and programmers etc., but I'm not. That's not what I mean. I like this project, and hope to continue being a part of it for some years if possible and need be :) :) )
Anyhow- thanks!!!
Eric

Perhaps copy and paste from boinc manager the postponed messages along with several lines before and after them to a reply to this topic. I may not be able to help, but someone else might be able to, and more information, in this case, would be better.

Eric,

Like geophi, I have no MAC experience. However, is space allowed for boinc projects sufficient? Free space erodes when CPDN tasks fail because CPDN is not good cleaning-up after failures. In Windows (and I think it was the same when I ran Linux years ago) tasks can fail for lack of space. Perhaps MAC OS interaction with boinc is more tolerant, hence the series of postponements. (Admittedly, that is pure conjecture on my part.)

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

Eric

Your computer has 2 processors, but your Tasks page shows that there are 4 models trying to run.

So, perhaps if you do this:
In the Projects tab, set cpdn to No new tasks.
In the Tasks tab, Suspend the 2 models that are least advanced.

This will allow just those 2 models to get on with it, instead of 4 of them swapping back and forth. When they finally finish and upload, you can Unsuspend the other 2 and run them.
(Excess tasks may be due to you having too large a queue.)

Some other things:
You only have 2 Gigs of memory. This is a bit low for climate models + the operating system + your work.
I looked at one of your failed models, and there's a lot of Suspends in the error list. This might be OK for other projects, but not for cpdn. Here you should set Suspend work if CPU usage is above to zero to turn it off.

Eric

Scouting around a bit more, it's possible that we're on the wrong track.
The "postponed" message appears to be related to a VM error, which you'd only get if you're running BOINC + VirtualBox.
VirtualBox is only needed if you're running one of the few projects that need it. e.g. the Cern stuff. It's definitely not need for this project.
There are stand-alone BOINC versions, which you'd be better off with unless you really need VirtualBox.