Re: SMC-Ultra ethernet sluggishness

Radovan Garabik (garabik@atlas02.dnp.fmph.uniba.sk)
Fri, 22 Oct 1999 17:20:53 +0200


Scott G. Miller wrote:
: >
:
:
: > I have an SMC-Ultra EtherEZ ISA adapter, which operates well most of the
: > time. However, after a lot of traffic (maybe a certain amount of data, or
: > a certain period of time), performance becomes abysmal. Connections to
: > anything anywhere stall for minutes at a time, or die out
: > entirely. Transfers, when they work, operate at a 1 or 2 k per second
: > rather than the usual 300k. This is not network performance in general.

the very same happens to me. (the same adapter)

: >
: > If I reset the network stack, by bringing eth0 down and removing the
: > modules (smc-ultra and 8390), then reinsert them, performance returns to
: > normal.

I do not need to reinsert modules; just bringing eth0 down and up returns it
to normal

: >
: > This is on Linux 2.2.13, on an ASUS SMP motherboard, with only one P2-300
: > processor (kernel is NOT compiled for SMP). I'm including
: > /proc/modules, /proc/interrupts and /proc/ioports just in case.

it goes all the time I remember (it was about 2.2.5 since the computer exists),
up to 2.2.12, I haven't tried 2.2.13 yet

:
: Useful information would also include the driver detection message,
: giving whether the card is operating in shared memory or I/O mode,
: the output of "cat /proc/net/dev" and any error/warnings reported
: by the driver (use "dmesg").

I/O mode, no errors or warning, everything seems usual

after 13 days of uptime, I think it did not happened since then
(it is not an important server anymore)

I vaguely remember that when it happened the number of errs went up
dramatically.

cat /proc/net/dev:
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 9890796 78049 0 0 0 0 0 0 9890796 78049 0 0 0 0 0 0
eth0:2115111477 10903678 10 0 0 310 0 6069402 111633103 5965476 32 0 0 60792 64 0
vmnet0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
vmnet1: 0 0 0 0 0 0 0 0 0 3183 0 0 0 0 0 0
vmnet2: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
vmnet3: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

:
: If there is nothing obviously wrong in that then you may need to use
: tcpdump to get a better look at what is happening during a slowdown.

nothing unusual, just a lot of traffic (i.e. copying files from nfs mount,
while ftp-ing (simultaneously more connections) other files, while
another linux with ipx-overbroadcasting bug was flooding the network :-)
(the latter is not neccessary)

-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__         garabik @ fmph.uniba.sk           |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!

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