[ENBD] ENBD problem

[Peter T. Breuer]

"A month of sundays ago Alexandru COSTIN wrote:"
>     On the server side I get this output after executing "/tmp/nbd-server
> 1090 /machines/remote  -t 1200" , when trying to connect with this command
> line "/tmp/nbd-client demo 1090 demo /dev/nda" (demo is the server's name)

I'll just interpret the log for you ...

> [root@eunet-demo nbd-2.4.14]# nbd-server: notice: setsockopt RCVTIMEO failed
> with Protocol not available

Linux problem. Nothing fatal. There follow messages indicating
scuccessful negotiation with the client over resource sizes and
so on. They agree one channel, as you specified.

> [root@eunet-demo nbd-2.4.14]# nbd-server: server (-1) opened port #1090 on socket 1
> nbd-server: server (-1) read passwd ok
> nbd-server: server (-1) got cliserv magic ok
> nbd-server: server (-1) sent size 1677721600 ok
> nbd-server: server (-1) sent sig ok
> nbd-server: server (-1) suggested ro flags 0 ok
> nbd-server: server (-1) received blksize 1024 ok
> nbd-server: server (-1) sent/negotiated blksize 1024 ok
> nbd-server: server (-1) agreed 1 channels ok
> nbd-server: server (-1) selected free port at 1092
> nbd-server: server (-1) posted port 1092 ok
> nbd-server: server (-1) manager started new process group 805
> nbd-server: server (-1) manager set CHLD USR1 USR2 HUP TERM signal handlers
> nbd-server: server (0) opened port #1092 on socket 6

The server then tells us that the client has closed the communication
socket:

> nbd-server: Read returned 0 after select predicted read!
> nbd-server: Read got 0 bytes in readnet: Invalid argument
> nbd-server: Not enough magic in packet. Breaking off.

The server decides that this is serious and breaks the session and
prepares to renegotiate from scratch on a new session. You can't talk
to a closed socket. The error will be at the other end.


> nbd-server: sighandler relaunches child from manager
> nbd-server: server (-1) main childminder checking pid 806

OK.  The managing daemon relaunches a daemon which will wait for
a connection, I think.

>     On the client side it says "client(-1) introduction sequence ends ok".

You need to look at the log messages too. Apparently the client has
closed the connection and is probably dead.

> I've read the alias thing and I've used it.
>     /proc/nbdinfo showsa-p Closed

Well, that's even more dead. It's never been alive. The client
apparently dies. I would guess that the module is extremely kaput.
Your client side logs might show all kinds of faults and oopses.
It's probably a module compiled for uniprocessor loaded into an smp
kernel, or vice versa.

>     Trying now to format the /dev/nda on the client machine ends
> unsuccesfully with a "Attempt to read a block from filesystem resulted in a

You are trying to format a device that doesn't exist. The kernel will
error or oops.

> short read while creating root dir" error. Mounting doesn't work, too.
> (Input/Output error)

It can't work. There's nothing to work on.

Peter