[R-SIG-Mac] Problems with big objects in R-2.5 x86_64

[Simon de Bernard]

Hi,

I don't know what your status is on your issue but I'd like to report  
that I managed to circumvent mine by using Doug Lea's malloc instead  
of the standard one.

Although R seems to be a bit slower (but not by much), at least it  
does not crash or go into infinite loops anymore when dealing with  
large objects.


Best,


Simon de Bernard.

Le 24 mai 07 à 10:15, Eric Blanc a écrit :

> Hi Simon,
>
> Apologies for not being precise enough.
>
> The problem I experience (R hanging while loading large objects past
> the 4GB mark) appears on a Mac Pro fitted with 16 GB of RAM, and
> after I had removed the process limits set by the shell (ulimit -d
> unlimited).
>
> Has anybody got the same experience as I had ?
>
> Thanks for your help (and of course for the efforts put in porting R
> and its great front end !)
> Eric
>
> On 23 May 2007, at 18:03, Simon Urbanek wrote:
>
>> Eric,
>>
>> unless you tell us more precisely what you do and what machine you
>> have we can't really help you. If you don't have enough RAM (>4GB)
>> 64-bit R won't really help you as it ends up swapping all the time
>> anyway (just look at the resource usage).
>>
>> Cheers,
>> Simon
>>
>>
>> On May 23, 2007, at 11:15 AM, Eric Blanc wrote:
>>
>>> Hi,
>>>
>>> I encountered problems in using the experimental build R 2.5 for
>>> x86_64 architecture. I have downloaded the framework from
>>>
>>> http://r.research.att.com/R-25-br-quad.tar.gz
>>>
>>> I tried to load large objects, (about 2GB each), but it hangs at
>>> random points, apparently always after the 4GB mark (i.e. while
>>> loading the 3rd object). I am monitoring the resources usage from
>>> another terminal window (top), and the R process never completes
>>> loading the 3rd object, the memory usage stops increasing before
>>> reaching 6GB, even though the CPU usage by R is 100%. The time
>>> required to load the 1st and 2nd objects are about 1 minute each,
>>> while I stopped R after 30 minutes (without completing loading the
>>> 3rd object). I had similar problems last december, when I tried to
>>> compile R 2.4 from scratch using the instructions in
>>>
>>> https://stat.ethz.ch/pipermail/r-sig-mac/2006-November/003384.html
>>>
>>> Here is the sessionInfo():
>>>
>>> R version 2.5.0 Patched (2007-05-02 r41420)
>>> x86_64-apple-darwin8.9.1
>>>
>>> locale:
>>> C
>>>
>>> attached base packages:
>>> [1] "stats"     "graphics"  "grDevices" "utils"     "datasets"
>>> "methods"
>>> [7] "base"
>>>
>>> Has anybody tried to use large amounts of memory with R 2.5 x86_64 ?
>>> I must add that I tried to load the same objects, on the same
>>> machine, but using Ubuntu 7.4 x86_64, and it works on Ubuntu. It
>>> seems to be a OS X problem (I am using 10.4.9).
>>>
>>> Thanks,
>>> Eric
>>> 	[[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> R-SIG-Mac mailing list
>>> R-SIG-Mac at stat.math.ethz.ch
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>>
>>>
>>
>
>
>
>
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-Mac mailing list
> R-SIG-Mac at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac