> I've just upgraded an FC5 system to FC7 test 4 and it hangs during boot
> when trying to start rpcbind. If I do an interactive boot and tell it
> not to start rpcbind then it'll get all the way to login quite happily,
> although anything that depends on rpc running obviously doesn't come up.
>
> Strangely, if I then start rpcbind, it'll come up quite happily.
>
> Is there anything obvious I ought to be checking, before I dive in and
> look more closely, or should I give up and do a clean install? (The
> purpose of this one was to practice the upgrade on a spare machine prior
> to playing with the main one)
>
> Dave
Has anyone else run into this? I just started to, for some reason. Fedora 7 had been working just fine. I set up NIS and suddenly rpcbind practically hangs the system. I say "practically" because the system will still run, but it takes 10 minutes to execute a simple directory listing ("ls -l"). BTW, the system load is nominal--nothing spiking the processor.
What I've found is that if I start rpcbind before network then everything cooks along perfectly. If I start network with no interfaces brought up (not even the loopback) it still causes rpcbind not to listen on any port. Another thing strange: rpcbind would complain that it couldn't bind (both tcp and udp) because the "address (*) was in use," but netstat doesn't show anything at all. I booted to single-user mode and just tested various scenarios so I know there are no other services masking the real problem.
I've switched the order of startup so that rpcbind starts before network, but I'd rather understand and fix the problem rather than hack around it.
Any ideas??
|
|||||||
-- fedora-list mailing list fedora-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
