"Matt" == Matt Dillon [EMAIL PROTECTED] writes:
[... my newfw bomb deleted ...]
Matt Heh heh. Yes, newfs has some overflows inside it when you
Matt get that big. Also, you'll probably run out of swap just
Matt newfs'ing the metadata, you need to use a larger block size,
Matt large -c
"Matt" == Matt Dillon [EMAIL PROTECTED] writes:
Matt Use the vn device to create a swap-backed filesystem. 'man
Matt vnconfig'. (In current VN functionality has been merged into MD
Matt and MD in swap-backed mode should be used instead ). If you
Matt turn softupdates on on the VN based
:Hmmm... I was just having a little fun, and I think that someone's
:using the wrong type of integer somewhere:
:
:[1:23:323]root@news:~ vnconfig -e -s labels -S 1t vn0
:[1:24:324]root@news:~ disklabel -r -w vn0 auto
:[1:25:325]root@news:~ newfs /dev/vn0c
:preposterous size -2147483648
:
:Dave.
On Sun, Feb 25, 2001 at 05:44:13PM -0500, David Gilbert wrote:
"Matt" == Matt Dillon [EMAIL PROTECTED] writes:
[... my newfw bomb deleted ...]
Matt I had a set of patches for newfs a year or two ago but never
Matt incorporated them. We'd have to do a run-through on newfs to
Matt
:Making the run for larger block sizes puts us in the same league as
:DOS. While it will stave off the wolves, it will only work for so
:long give Moore's law.
:
:Dave.
:
:--
:
:|David Gilbert, Velocet Communications.
:With 512 Byte blocksizes you are limited to 1T because the physical
:block number is a signed 32bit.
:FFS uses 32bit (I wouldn't count on the high bit) frag numbers.
:A fragment defaults to 1k so even with 1k fragments the limit is
:at least 2T.
Yes, the FFS limit is essentially the frag
We have an application which is precompiled, for linux, and stupid.
It uses (at times large) scratch files. We want to run this on our
diskless machines (CPU farm) to cut the per-cpu cost of the
computation ($200/drive starts getting significant at a few hundred
machines).
Anyways, I've tried:
On Sat, Feb 24, 2001 at 04:08:04PM -0500, David Gilbert wrote:
We have an application which is precompiled, for linux, and stupid.
It uses (at times large) scratch files. We want to run this on our
diskless machines (CPU farm) to cut the per-cpu cost of the
computation ($200/drive starts
"Bernd" == Bernd Walter [EMAIL PROTECTED] writes:
Bernd You have to raise MAXDSIZE and DFLDSIZ in your kernel config.
Bernd You may also want to try MD instead of MFS.
Well... I was reading about md and the man page seemed to imply that
it was _very_ dependant on what malloc(9) could do ...
:On Sat, Feb 24, 2001 at 04:08:04PM -0500, David Gilbert wrote:
: We have an application which is precompiled, for linux, and stupid.
: It uses (at times large) scratch files. We want to run this on our
: diskless machines (CPU farm) to cut the per-cpu cost of the
: computation ($200/drive
10 matches
Mail list logo