tulip.o 2.2.15, 2.2.16 shaky performance

From: Jens Benecke (jens@pinguin.conetix.de)
Date: Wed Aug 02 2000 - 04:43:01 EST

Hi all,

the following happens here on two K6-(500|350) machines.

NFS Server: 2.2.15-6mdk
        K62-350, 64MB RAM, IDE UDMA33 with two IBM disks
        knfsd, version 1:

NFS Client: 2.2.16-9mdk
        K62-500, 128MB RAM, SCSI (ncr53c8xx, Tekram DC-390F)
        nfs-utils-clients.rpm version 0.1.5

Copy files (on client) from NFS volume to local /tmp: about 3-3.5MB/s,
steady throughput (according to xosview).

Copy (on client) from local harddisk to remote dir: very shaky, 1sec
throughput and 2sec pause, avg. about 200-500k/sec. This happens also when
I replace the network card in the client (with a RTL 8139).

Another issue is that whenever I unplug the tulip card from the network
switch, it stops working until I unload and reload the tulip module.
(old_tulip does not help, does the same thing). The rtl8139 module can
apparently cope with this without reloading.

I cannot just start experimenting on the server because I'd have to explain
the downtime to 40 mad scientists =;) so if this is a known problem I'd
appreciate if someone could point me in the right direction.

Please CC me, just in case I miss something in the list. Thanks. :)

ciao, Jens (mailaddr im Header)                 http://www.pinguin.conetix.de
"Schiebe nie etwas auf Boshaftigkeit, was           http://www.hitch-hiker.de
ausreichend durch Dummheit erklärt werden kann."       http://www.linuxfaq.de

- 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/

This archive was generated by hypermail 2b29 : Mon Aug 07 2000 - 21:00:08 EST