This problem occurs in the default RedHat6 kernel, the updated RedHat6
kernel,
the vanilla 2.2.9 and 2.2.7-ac1, 2.2.7-ac4
Here is the issure I am having, whenever the i do a smbclient -L
hostname for this particular server I get the following message in the
logs:
open_oplock_ipc: Failed to get local UDP socket for address 100007f.
Error was Cannot assign requested address
that is from the smb.log that gets repeated mulitple times after a few
seconds i get this message in the nmb.log:
[1999/06/08 16:02:39, 0] libsmb/nmblib.c:send_udp(725)
Packet send failed to 0.0.0.0(138) ERRNO=No buffer space available
I also see this in the boot messages when samba starts:
kernel: neighbour table overflow
also if I try to ping the address of the host i get these error
messages:
ping: sendto: No buffer space available
ping: wrote 192.168.0.1 64 chars, ret=-1
another symptom i belive is that when i am at home connected to the
internet by ppp, i can ssh/telnet into this box, but if i do an 'ls'
the Xterm just freezes, i do a netstat on the machine i'm ssh'd into and
it shows lots of packets in the send-Q
and my machine at home show packets in the Recv-Q, but the output never
gets displayed, this problem does not show itself if i go into the
machine from another computer on the lan. Also apache will stall across
the ppp connection but not the lan connection.
I added the /dev/pts entry into /etc/fstab as I was getting error
messages about no buffer space available when i tried to ssh/telnet into
the box, Is there another device entry I am missing?? Is their some
sort of kernel tuning i am not aware of?? These problems seem to
manifest after the upgrade to RedHat6 but they did not seem to manifest
themselves immediately.
I am running alot of network services on this machine,
X,ssh,apache,ftp,samba.
Any assistance is appreciated, I would gladly provide more detailed
info, I'm just not sure what to provide.
Trey Hardin
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
