I've been watching the memory usage of boinc.exe increase slowly from start up. Currently at 68MB, started below 20MB (I think). I've seen it as high as 600MB before a computer crash. Observations are with 6.10.50/.51 and now .45.

Win7 64bit. 2GB DDR3 RAM. Q9450. HD5970 & HD4870.

Projects:
Aqua
Collatz
FreeHAL
LHC
Milkyway
MalariaControl
QMC

I just cross-posted this on the BOINC Alpha list ...

I just cross-posted this on the BOINC Alpha list ...

Thanks Paul, for some reason I can't email that list. Up to 88MB now.

I just cross-posted this on the BOINC Alpha list ...

Thanks Paul, for some reason I can't email that list. Up to 88MB now.

Yeah, well, not likely to do any good... they ignore most everyone now these days ... heck, JM VII can't seem to get a hearing anymore ... RIchard is still running about 60/40, with the bulk ignored ...

A new poster pointed to a problem with work fetch and idling resources ... I pointed out that this is just one of many lingering issues (though I did not point out that I had previously posted about idling resources because there was not a proactive enough work fetch) and suggested a more comprehensive overhaul after a review of the prior year's posts on the subject...

So far, deafening silence ...

The real problem is that the internal "model" is still based on a single processing unit running at a fairly slow rate of speed... this pervades the internal structures and the "thinking" behind the code... all of the features added are more or less bolt ons to this model resulting in a messy kludge ... that it works at all is amazing...

I also started to monitor one of my machines to see if I can see the same thing ... and it looks like I am ... I started at 54,012 and right now, several hours later I see it at 60,784 ...

{edit - add}
posted an update with my and your growth numbers ...

{edit-add2}
On alpha list:
====================
Ah, thanks ...

Need someone to run a copy? Or will this be in the next drop?

If'n you do, both Paul (TGG) and I are running WIn7-64 and if you send me a copy I can proly get him to sign up as well to run a special drop ...

On May 11, 2010, at 12:54 PM, Rom Walton wrote:

We are looking into it, I've commited a bunch of changes to re-enable
the memory leak detection code.

Just to confirm your observations, my boinc is currently at 139MB! I should probably restart boinc.

This is version 6.10.43. Since I run CPU only, I haven't downgraded after the recall, although now I think I should.

I've been watching the memory usage of boinc.exe increase slowly from start up. Currently at 68MB, started below 20MB (I think). I've seen it as high as 600MB before a computer crash. Observations are with 6.10.50/.51 and now .45.

Win7 64bit. 2GB DDR3 RAM. Q9450. HD5970 & HD4870.

Sounds like a good Save. Saw this and turned on MW to see what happens, had been running DNETC all night so far, and boinc.exe was at 8,804K.

At the start it went to just over 10,000K. It stayed pretty steady during the WU run +/-s but not much. At the end of the crunch when new WUs loaded, it jumps by between 50 & 150K. At present its at 11,216K after around 10/15 mins or so of MW crunching. For whatever reason it did not appear to change when crunching DNETC remaining at 8,804K all evening, no dispersion cast, pure observation.

Looks like the leak suspicion is gaining traction, it is steadily climbing, hugely no, but it never goes down either, so over time a big number will certainly build as illustrated above. My environment:

Phenom2 940 o/c 3.2Ghx 8Gb RAM, 5970, Aqua on CPU, previously DNETC all night GPU, short 10/15 mins test MW GPU, BOINC 6.10.45, ATI 10.3

EDIT: Still on the up, 11,365K at present - averages about +50K each time a WU run finishes, and new ones loaded. Only significantly jumps when WUs are changing over.

Regards
Zy

Rom says he thinks he has found the leak and the next drop should contain the fix for it ... which would be 6.10.54 ...

Rom says he thinks he has found the leak and the next drop should contain the fix for it ... which would be 6.10.54 ...

It may well have been fixed... but .54 itself is bad ... does not seem to start GPU work and freezes the system when a GPU is installed... so, don't try .54 ... or only try it on systems you don't want BOINC to run on ...

Thanks Paul.

I've installed the even newer .55. Starting memory is at 12.7MB. I'll be keeping an eye on both the memory usage and if a GPU goes idle. Thanks for your efforts with the alpha mailing list.

Live long and BOINC!

Varying between 13.1 and 13.3MB atm.

I use an old version of BM - 6.10.13 - and the memory use by boinc.exe has varied in the range 13.1Mb to 17.8Mb (as it is now).

My memory hungry WUs are those from Einstein, where each one of 4 is shown to use 197Mb.

---

Go away, I was asleep

Down to 7MB.

So, maybe they did find it... :)

I am glad that this build seems to have fixed your issue ...

WOn't know for sure till tomorrow, but it looks like my idle the GPU issue may also be fixed... though I am having an upload problem to SaH at the moment ... other projects seem to be fine ...

Just installed it (6.10.55) on the Mac as well, so 2 out of 5 machines over to the latest ...

Still hovering around 7MB. No idle GPUs either. Haven't tried any of the 'new' features though, but at least the basics appear to be working.

Still hovering around 7MB. No idle GPUs either. Haven't tried any of the 'new' features though, but at least the basics appear to be working.

Yeah ...

I lost 15 hours of GPU compute time on one system because I got hit with the GPU Idle bug and I missed checking on the system soon enough ... was runnign 6.10.45 which is relatively immune though not completely ... On the one WIn7 machine I have it running on 12 hours in and still going ... usually I would be toast by now ...

I noticed a couple minor issues with debug messages I posted ... but, .55 looks like it will be my fall-back to replace .45 unless something more subtle and evil comes up ...