eth0: Memory squeeze, deferring packet

[Jason Lim]

Hi!

Do you'all know what this means:

"eth0: Memory squeeze, deferring packet"?

We get that one one of our boxes every so often, and it is annoying
because then it loses all connectivity to the net, and has to be
physically rebooted.

eth0: RealTek RTL8139 Fast Ethernet at 0xe800, IRQ 10, 00:50:fc:28:46:90.

So... anyone know whats going on, or experienced this before?

Any help would be greatly appreciated.

Sincerely,
Jason

--
To UNSUBSCRIBE, email to debian-is...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org

[Jason Lim]

Hi,

Thanks for the reply.

Strange... but wha could be causing it? The box has 512M ram and has
almost 20M free... it is that empty and free.

Plus there is no huge amount of traffic going through it.

I'm just trying to figure out what triggered this :-/

Sincerely,
Jason

----- Original Message -----
From: "Russell Coker" <rus...@coker.com.au>
To: "Jason Lim" <mail...@jasonlim.com>; <debia...@lists.debian.org>
Sent: Sunday, October 07, 2001 2:06 AM
Subject: Re: eth0: Memory squeeze, deferring packet

> On Sat, 6 Oct 2001 19:12, Jason Lim wrote:
> > Memory squeeze, deferring packet
>
> The dev_alloc_skb() function in the kernel is failing due to lack of
memory.
>
> Presumably if you change /proc/sys/vm/freepages to have some larger
numbers
> it will stop this happening.
>
> --
> http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
> http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
> http://www.coker.com.au/projects.html Projects I am working on
> http://www.coker.com.au/~russell/ My home page

[Russell Coker]

On Sun, 7 Oct 2001 02:41, Jason Lim wrote:
> Strange... but wha could be causing it? The box has 512M ram and has
> almost 20M free... it is that empty and free.
>
> Plus there is no huge amount of traffic going through it.
>
> I'm just trying to figure out what triggered this :-/

These types of things are caused by one of two situations:
1) Some situation of extremely high system load (eg a combination of intense
file system access over software RAID and routing between several 100baseT
interfaces).
2) Kernel bug.

If it's the latter than having 2G of RAM isn't necessarily going to save
you...

--
http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
http://www.coker.com.au/projects.html Projects I am working on
http://www.coker.com.au/~russell/ My home page

[Jason Lim]

Well... I doubt the it is (1) because the system was NOT under great load,
and doesn't have software RAID, and only uses one 100baseT realtek card.

SO... i guess it is (2). Kernel is 2.2.19... i thought it would be pretty
much stable by now?! damn.

how is kernel 2.4.x now? Do you think they are safe enough for production
use? Any huge performance increase?

Sincerely,
Jason

----- Original Message -----
From: "Russell Coker" <rus...@coker.com.au>
To: "Jason Lim" <mail...@jasonlim.com>; <debia...@lists.debian.org>

Sent: Sunday, October 07, 2001 7:35 AM
Subject: Re: eth0: Memory squeeze, deferring packet

> On Sun, 7 Oct 2001 02:41, Jason Lim wrote:
> > Strange... but wha could be causing it? The box has 512M ram and has
> > almost 20M free... it is that empty and free.
> >
> > Plus there is no huge amount of traffic going through it.
> >
> > I'm just trying to figure out what triggered this :-/
>
> These types of things are caused by one of two situations:
> 1) Some situation of extremely high system load (eg a combination of
intense
> file system access over software RAID and routing between several
100baseT
> interfaces).
> 2) Kernel bug.
>
> If it's the latter than having 2G of RAM isn't necessarily going to save
> you...
>
> --
> http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
> http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
> http://www.coker.com.au/projects.html Projects I am working on
> http://www.coker.com.au/~russell/ My home page
>
>

[Russell Coker]

On Sun, 7 Oct 2001 06:13, Jason Lim wrote:
> Well... I doubt the it is (1) because the system was NOT under great load,
> and doesn't have software RAID, and only uses one 100baseT realtek card.
>
> SO... i guess it is (2). Kernel is 2.2.19... i thought it would be pretty
> much stable by now?! damn.
>
> how is kernel 2.4.x now? Do you think they are safe enough for production
> use? Any huge performance increase?

2.4.x is working well on my servers. It is supposedly faster at IP routing
and certain types of disk IO. My tests show better raw ide performance but I
never seriously compared file system speed.

--
http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
http://www.coker.com.au/projects.html Projects I am working on
http://www.coker.com.au/~russell/ My home page