Re: Machine dies under heavy I/O or network-access ..?

[Christiaan den Besten]

Hi !

Our situation has cleared up a little bit. Our Areca Raid-controller had troubles with Maxtor SATA disks and died eventually. They 
fixed this in a new firmware release (at least, we are testing that right now ...).

I have been playing with sysctl.conf as well:

---
# 5s will give us between 200 and 260Mb dirty buffer ...
vm.dirty_background_ratio = 5
vm.dirty_ratio = 75
---

By keeping the dirty buffer somewhat lower (250M) it seems to be preventing the system from using too much memory. If I recall 
correct I have seen 'some' assertions, but never fatal anymore.

bye,
Chris

----- Original Message ----- 
From: "Pauli Borodulin" <pauli.borodulin@xxxxxx>
To: "Christiaan den Besten" <chris@xxxxxxxxxxx>
Cc: <linux-kernel@xxxxxxxxxxxxxxx>
Sent: Thursday, October 13, 2005 12:44 PM
Subject: Re: Machine dies under heavy I/O or network-access ..?

Christiaan den Besten wrote:
> [...]
> Today I noticed the following assertion in dmesg:
> ---
> e1000: eth0: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex
> KERNEL: assertion (!sk->sk_forward_alloc) failed at net/core/stream.c (279)
> KERNEL: assertion (!sk->sk_forward_alloc) failed at net/ipv4/af_inet.c
> (148)
> ---
> [...]

I upgraded ten of our servers to 2.6.13.3 on tuesday and noticed today
that the following assertations were in one server's dmesg:

KERNEL: assertion (!sk->sk_forward_alloc) failed at net/core/stream.c (279)
KERNEL: assertion (!sk->sk_forward_alloc) failed at net/ipv4/af_inet.c (151)

This server has e1000 NIC just like most of the servers I upgraded. The
difference in our situation is that the server which logged the
assertions is still running fine. Previous kernel version on the servers
was 2.6.12.2 and it was used over 90 days without any problems.


Br,
--
Pauli Borodulin <pauli.borodulin@xxxxxx>
Systems Analyst, tel. +358 3 3551 7892
Computer Centre / Room B4179
University of Tampere, Finland 

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