From: Zed Pobre <zed@resonant.org>
To: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: Complete freeze with 2.4.20 on 4-proc IBM xSeries 350
Date: Thu, 26 Dec 2002 15:43:38 -0600 [thread overview]
Message-ID: <20021226214338.GA1285@resonant.org> (raw)
In-Reply-To: <11930000.1039565421@flay>
[-- Attachment #1: Type: text/plain, Size: 6428 bytes --]
On Tue, Dec 10, 2002 at 04:10:21PM -0800, Martin J. Bligh wrote:
> > I'm experiencing serious problems with Kernel 2.4.20 on a IBM xSeries 350
> > machine, having 4 700 MHz processors and 4 GB RAM (same on another machine
> > with the same configuration, but only 3 GB RAM). The machine just com-
> > pletely freezes after some time, ranging from 20 minutes to 3 hours. It
> > is running IBM DB/2 with quite some load, the base system is RedHat 7.2
> > with all the updates applied. There is no oops or other fault, just a
> > plain freeze.
>
> Can you watch /proc/meminfo and see how low "lowfree" gets?
> If it gets low (eg below 50Mb), dump /proc/slabinfo as well.
I'm running off a Soyo Dragon KT400 board, single AMD 2400+
processor, 1GB ram (4GB limit set in kernel), and was also having
freezes, apparently related to having software raid 1 in use (on two
drives sitting on the on-board Highpoint controller). If I started a
large data transfer over the net, the machine would freeze after some
time ranging from a few minutes to an hour, with either 2.4.20-ac2 or
2.4.21-pre2. I tried letting the system resync before restarting, and
the first pass through it actually froze during resyncing, with
nothing special going on. Later, after a resync, I started trying all
sorts of things, from turning off XFree86 and running a minimal set of
loaded modules, to using mdadm to fail one of the drives (it was a
RAID1 setup) to force only a single disk to be in use. Right now,
I've taken it off of RAID entirely, and am mounting /dev/hdex
directly, and copying data to that, and it seems to be okay so far
(for the last hour or so).
cat /proc/meminfo gives:
total: used: free: shared: buffers: cached:
Mem: 1055412224 1045782528 9629696 0 90771456 720228352
Swap: 526376960 39202816 487174144
MemTotal: 1030676 kB
MemFree: 9404 kB
MemShared: 0 kB
Buffers: 88644 kB
Cached: 684644 kB
SwapCached: 18704 kB
Active: 540556 kB
Inact_dirty: 356560 kB
Inact_clean: 44764 kB
Inact_target: 188376 kB
HighTotal: 131008 kB
HighFree: 1024 kB
LowTotal: 899668 kB
LowFree: 8380 kB
SwapTotal: 514040 kB
SwapFree: 475756 kB
Committed_AS: 322008 kB
So if you were concerned about LowFree going low, it is.
/proc/slabinfo:
slabinfo - version: 1.1
kmem_cache 68 68 112 2 2 1
fib6_nodes 5 112 32 1 1 1
ip6_dst_cache 5 20 192 1 1 1
ndisc_cache 1 30 128 1 1 1
ip_conntrack 10 24 320 2 2 1
tcp_tw_bucket 0 0 128 0 0 1
tcp_bind_bucket 25 112 32 1 1 1
tcp_open_request 0 0 128 0 0 1
inet_peer_cache 0 0 64 0 0 1
ip_fib_hash 9 112 32 1 1 1
ip_dst_cache 12 40 192 2 2 1
arp_cache 3 30 128 1 1 1
urb_priv 4 59 64 1 1 1
blkdev_requests 1408 1410 128 47 47 1
nfs_write_data 0 0 384 0 0 1
nfs_read_data 0 0 384 0 0 1
nfs_page 0 0 128 0 0 1
journal_head 26 77 48 1 1 1
revoke_table 1 250 12 1 1 1
revoke_record 0 0 32 0 0 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 4 40 96 1 1 1
fasync_cache 1 202 16 1 1 1
uid_cache 6 112 32 1 1 1
skbuff_head_cache 163 260 192 13 13 1
sock 201 204 896 51 51 1
sigqueue 1 29 132 1 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 22 472 64 8 8 1
bdev_cache 9 59 64 1 1 1
mnt_cache 18 59 64 1 1 1
inode_cache 54639 57071 512 8153 8153 1
dentry_cache 804 4050 128 135 135 1
dquot 0 0 128 0 0 1
filp 1563 1590 128 53 53 1
names_cache 0 2 4096 0 2 1
buffer_head 185515 191520 128 6368 6384 1
mm_struct 63 80 192 4 4 1
vm_area_struct 2749 2850 128 92 95 1
fs_cache 62 118 64 2 2 1
files_cache 62 63 448 7 7 1
signal_act 71 75 1344 24 25 1
pte_chain 70104 74592 8 221 222 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 2 2 65536 2 2 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 2 2 32768 2 2 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 2 7 16384 2 7 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 15 19 8192 15 19 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 121 139 4096 121 139 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 50 130 2048 26 65 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 53 60 1024 14 15 1
size-512(DMA) 0 0 512 0 0 1
size-512 99 120 512 13 15 1
size-256(DMA) 0 0 256 0 0 1
size-256 86 120 256 6 8 1
size-128(DMA) 3 30 128 1 1 1
size-128 953 990 128 33 33 1
size-64(DMA) 0 0 64 0 0 1
size-64 776 1062 64 17 18 1
size-32(DMA) 52 59 64 1 1 1
size-32 714 767 64 13 13 1
Does this provide you with any useful information?
--
Zed Pobre <zed@debian.org> a.k.a. Zed Pobre <zed@resonant.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.
[-- Attachment #2: Type: application/pgp-signature, Size: 481 bytes --]
next prev parent reply other threads:[~2002-12-26 21:36 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-12-10 22:58 Complete freeze with 2.4.20 on 4-proc IBM xSeries 350 "Rüegg, Peter H."
2002-12-11 0:10 ` Martin J. Bligh
2002-12-26 21:43 ` Zed Pobre [this message]
2002-12-28 4:28 ` Zed Pobre
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20021226214338.GA1285@resonant.org \
--to=zed@resonant.org \
--cc=linux-kernel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.