where is k_compat.h?

[Paul Edwardson]

Here are answers to James' questions regarding my 3c905b-tx-nm stalling:

> Moderately heavy traffic to where?  To a machine on the same switch,
> elsewhere on your LAN, or elsewhere in the world?
> 
> What types of connections slow to a creep and then stall?  TCP/IP
> connections, all IP connections, or all connections of any type?

I create the problem by displaying a series of simple gnuplot
x-windows one per second (which isn't really heavy traffic at all) on
any of several machines connected to either the same 100 Mbps, another
100 Mbps, or a 10 Mbps switch. An interesting thing is one of the
machines on another 100 Mbps switch is a 450 Mhz PIII with another
3c905b-tx-nm NIC but purchased several months ago. That machine works
great.

Does that suggest this could be a problem somewhere other than in the
NIC or kernel? By the way, the older machine that works is running
Mandrake 2.2.13-4mdk and the newer one that fails does so with that
same kernel or with 2.2.13-22mdk.

The problem also occurs during "mount -t nfs -a" of 9 disks on
various machines.

If I ping from the stalling machine to another, or vice versa, the
packet loss rate is around 20% and those that survive report times
about 10000 times normal. Eventually the loss rate becomes 100%. Have
I given enough info to answer your question about which kinds of
connections slow? If not, let me know what other tests I need do to
answer it.

>    Try turning off the RFC1323/RFC2018 features of the Linux 2.2.*
>    kernel, by doing the following:
>
>        # cd /proc/sys/net/ipv4
>        # echo 0 >tcp_sack
>        # echo 0 >tcp_timestamps
>        # echo 0 >tcp_window_scaling
>
>    I suspect doing so will prevent the connection from stalling
>    entirely, but the throughput will probably still slow to a crawl.

Your suspicion was correct. This does prevent the complete stalling,
but the slowing does occur. The ping statistics that I described above
hold in this case too. Roughly 20% packet loss with times 10000 times
normal.

>    Try forcing both the switch and the NIC to half duplex mode.  In
>    my limited testing, I've found that if I force the NIC of the
>    machine that is *sending* the data to half duplex, the performance
>    problems mostly vanish.

I'll do this next and let you know what happens.

Thanks for your response.
Paul


-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-vortex-bug-request@beowulf.org