>> I'm using nfs to mount from a HP/UX server. That gives the "usual"
>> trouble, which I got rid of (more or less) by applying a patch by Bill
>> Hawes.
>> - But I ran in some weirdness, one of them being:
>> Mar 2 15:30:20 hyper kernel: nfs_revalidate_inode: eijk/.Xauthority getattr failed, ino=19455, error=-116
>>
>I started having trouble with 2.1.88 against a Linux NFS server.
>Processes on the server stopped being able to obtain authorization! When
>I rebooted the client machine, it complained bitterly during shutdown and
>spat out the same type of message you mention above. Afterwards,
>everything on the client began working again.. Looks like something had a
>lock on .Xauthority?
It happens on and off over here. And it disappears for no apparent
reason.
Mar 3 11:16:11 hyper kernel: nfs_revalidate_inode: eijk/.Xauthority
getattr failed, ino=19268, error=-116
Mar 3 11:16:11 hyper kernel: nfs_revalidate_inode: eijk/.Xauthority
getattr failed, ino=19268, error=-116
What's worse, the HP/UX patchs fixes more or less the trouble I'm
having with the truncated readdir requests to the HP, but I can no
longer reliably mount from another Linux box anymore, indifferent
which nfs server I use (user of kernel). This is what happens on
client side:
Mar 3 12:48:07 tommie kernel: RPC: task timed out
Mar 3 12:48:07 tommie kernel: nfs: server hyper not responding, still
trying
and the server is complaining about the following:
Mar 3 12:47:40 hyper kernel: UDP: bad checksum. From 8259B579:797 to 8259B567:2049 ulen 204
Mar 3 12:48:09 hyper last message repeated 9 times
I think this is due to the HP/UX fix, which fixes it for all servers
I've mounted, unfortunately also for the non HP servers.
Grtz
EJ
-- +--------------------+ There's only one rule: | Erwin J. van Eijk | The golden rule. | eijk@acm.org | He who owns the gold, rules. +--------------------+- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu