lkml.org 
[lkml]   [1999]   [May]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: Adding more than 8 swap partitions
Date
Boy...did I sure start something here...

"Dr . David Alan Gilbert" wrote:

> At 21:05 24/05/99 -0400, you wrote:
> >That May be true, but never the less you are now nearing the 2gig/4gig
> >barriers where signed ints, and unsigned ints may unexpectedly shaft u.
>
> My experience so far on the Alpha is that the size of int problems cause
> problems on all programs - not just those that might use a lot of RAM - so
> I don't
> think it becomes much more of an issue.
>

Well, I can say one thing - adding that much swap isn't the problem. I
successfully added ~5GB of swap earlier today before I decided that I was
satisfied that wasn't the problem I was seeing (not being able to add more than
X swap partitions). I just split up our 4GB disks into two sections (mkswap only
supports 2GB swap areas right now...unless I have an old mkswap). I swapped them
both on, as well as a few others, and voila!, 5+GB of swap. :)

>
> >But i'd love to see a 2/4gig core dump ! what will gdb say!. How long will it
> >take to dump :-)
>
> But that leads to the obvious question of how much RAM do you need to run GDB
> with a 2/4 GB core :-)
>

Ugh. One of the guys that I work with had the pleasure of watching (or
rather, not watching if you know what I mean) an 8GB core dump on an AS8400. I
don't remember how long it took, but suffice it to say that going out and eating
lunch wasn't long enough. ;)

<snip>

> Dave
>
> P.S. The other good argument for vast memory is multiprocessing - if the Alpha
> quad 21264 boards are around then it wouldn't be unreasonable to start
> thinking
> about 512MB-1GB of RAM for each processor.

<evil grin> I'm still trying to steal time on one of our ES40s...now _that_
will be fun. (4 CPU, 4 GB of memory, 20 some-odd disks - my kind of workstation
;)

As far as my earlier problems trying to mount >15 partitions and adding > 8
swap partitions, well, they seem to have cleared up (fingers crossed, knocking on
wood ;). I had to fiddle past a compile problem with the patch that Richard
Henderson sent out to the axp-list (does "alignment requested is not a constant"
in include/asm-alpha/smp.h at line 18 sound familiar to anyone?) but now that I
have that patched onto 2.2.9 and have changed MAX_SWAPFILES to 32 instead of 8, I
now have a system with 26 work-disks mounted (for a total of 30) and 27 swap
partitions turned on. I'm going to copy over the kernel to the Durango I'm
working on first thing tomorrow morning and see if it works there as well. As it
stands now, I've currently got a 2CPU Rawhide chugging away at an AIM VII shared
run. Hopefully everything will run nice and smoothly... Thanks to everyone who
sent in suggestions!

- Pete


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:51    [W:0.074 / U:0.384 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site