NFS screwiness

[Cloudmaster <sauer@cloudmaster.com>]

Does anyone off the top of tehir head know what would be causing these
messages?  I seem to be getting them a lot...  I know part of it is related
to the IDE crappiness in pyro's NFS server (but the disks cost *so* much
less than SCSI), but they seem too frequent.

Dec  6 19:30:12 cloud233 kernel: nfs: task 28691 can't get a request slot
Dec  6 19:30:12 cloud233 kernel: nfs: task 28692 can't get a request slot
Dec  6 19:30:12 cloud233 kernel: nfs: task 28635 can't get a request slot
Dec  6 19:30:12 cloud233 kernel: nfs: task 28694 can't get a request slot
Dec  6 19:30:12 cloud233 kernel: nfs: task 28695 can't get a request slot
Dec  6 19:30:13 cloud233 kernel: nfs: task 28697 can't get a request slot
Dec  6 19:30:17 cloud233 kernel: nfs: server pyro OK

Also, I get a lot of

Dec  1 14:47:12 pyro kernel: eth0: RTL8139 Interrupt line blocked, status 4.
Dec  1 14:47:12 pyro kernel: eth0: SMP simultaneous entry of an interrupt
handler.

Has anyone seen any mention of RTL8139 (cheap 10/100 ethernet card) problems
with SMP on any discussions?  It's done this with a few recent kernels, and
I think it may be causing me connectivity problems.  I'm thinking about
swapping it out for an Intel 10/100 card - and super success stories with
those?

That's enough bandwidth wasting for now...

--Danny


--
To unsubscribe, send email to majordomo@luci.org with
"unsubscribe luci-discuss" in the body.