tcp_bind: bad state, -4

[Ian Lochray]

I keep getting the error message "tcp_bind: bad state, -4" on my console. I think it is coming from rpc_bind but could not swear to this. DOes anyone know what is causing this and how to stop it either from occurring or from appearing on the console. There are no errors coming from any of the applications running on the server.

[Tom Geudens]

Hi,
#rpcbind -w
might do the trick ...

Regards,
Tom Geudens

[Steven Gillard_2]

I've never seen this before, but a couple of things to check:

- Does the message appear in dmesg? If so its coming from the kernel. I would ask HP about it.

- Does the message appear in syslog? Maybe you can get a process id from here.

Otherwise, you could try running lsof to find out if any processes have /dev/console open.

Regards,
Steve

[Steven Gillard_2]

hmmm, looks like it might be coming from the kernel after all:

# strings /usr/conf/lib/libinet.a | grep tcp_bind
tcp_bind
tcp_bindi
tcp_bind: bad ioctl, no b_cont
tcp_bind: bad req, len %d
tcp_bind: bad state, %d <<<<--------------
tcp_bind: bad address parameter, offset %d, len %d
tcp_bind: bad address length, %d
tcp_bind: no priv for port %d
tcp_bind: addr in use, new addr 0x%x, port 0x%x%s%s, old addr 0x%x, port 0x%x%s%s
tcp_bind: out of ports
tcp_bind
tcp_bindi

You might need to call HP on this one...

Regards,
Steve