kernel: nfs: task xxxxx can't get a request slot
Hello list,
I have two woody boxes, one serves as nfs server, the other as client.
They both link to the same switch. The server exports /var/www for the
nfs client mount. However, from time to time I receive the messages from
the nfs client:
nfs: server shmi OK
nfs: server shmi not responding, still trying
nfs: server shmi OK
nfs: server shmi not responding, still trying
nfs: server shmi OK
nfs: server shmi not responding, still trying
nfs: server shmi OK
nfs: server shmi not responding, still trying
nfs: server shmi not responding, still trying
nfs: task 48158 can't get a request slot
nfs: task 48159 can't get a request slot
nfs: task 48160 can't get a request slot
nfs: task 48161 can't get a request slot
nfs: task 48162 can't get a request slot
nfs: task 48163 can't get a request slot
nfs: server shmi OK
nfs: server shmi OK
nfs: server shmi OK
nfs: server shmi OK
nfs: server shmi OK
nfs: server shmi OK
nfs: server shmi OK
nfs: server shmi OK
Then I go to the console of the nfs server and it does not work and
prompt. Hence I have nothing to do but poweroff the nfs server.
I read the NFS faq at
http://nfs.sourceforge.net/
It says 3 reasons could possible lead into the error:
- Network congestion
- Overloaded server
- Packets (input or output) dropped by a bad NIC or driver...
I think it shouldn't be network congestion, for the two machines are
linked to the same switch. I am not sure the nfs server is overloaded.
If I can login the nfs server, the average load is usually 0.00. And,
I use kernel 2.4.18 with bonding support with 2 Intel ExpressPro 100 NIC
devices.
Any suggestion appreciated.
--
Patrick Hsieh <pahud@pahud.net>
Reply to: