samba - Nov 2000 - smb errors

Hi All,

	Welp, upgrading samba to 2.0.6 and using a 2.2.18pre21 kernel
does not seem to correct the problem when using smbmount to mount a
NT server share (here is the error message from /var/log/messages):

Nov 23 11:05:46 htmlodds kernel: smb_trans2_request: result=-32, setting
invalid
Nov 23 11:06:14 htmlodds kernel: smb_retry: signal failed, error=-3

df -h <enter> gives:

/dev/hda1             6.7G  1.2G  5.1G  20% /
NEW_DBS:SYS           8.0G  5.1G  2.9G  63% /.nwamd/NEW_DBS_SYS
df: /mnt/server1: Input/output error

an ls -al /mnt/server1 yields:

ls: /mnt/server1: Input/output error

now in /etc/hosts I have:

xxx.yyy.zzz.2 inet_server_1 server1

and the command I used for smbmount is:

smbmount \\\\inet_server_1\\C$ /mnt/server1 -U <username>

now, the only difference is this server is not plugged into the same switch
as inet_server_1 and the other three linux boxes...could this make a
difference as to why this is happening?

I have to be able to access files on this NT Server share (for reading
only) if this conversion to linux is going to be successful.

-Bill

On Thu, 23 Nov 2000, Bill Parker wrote:
> Hi All,
> 
> 	Welp, upgrading samba to 2.0.6 and using a 2.2.18pre21 kernel
> does not seem to correct the problem when using smbmount to mount a
> NT server share (here is the error message from /var/log/messages):
You are not describing the problem very accurately, not everyone mounting
something from an NT Server is seeing this (the problem vs my problem).
> Nov 23 11:05:46 htmlodds kernel: smb_trans2_request: result=-32, setting
> invalid
> Nov 23 11:06:14 htmlodds kernel: smb_retry: signal failed, error=-3
But this says that you got the mount "up" (I think), then there was an
error and refreshing the mount failed. It does not say that the mount
itself failed.
This was at 11:05:46, how much before that did you mount? Seconds?

For some reason the smbmount (mount.smbfs) process is no longer alive
(that's what the -3 says).

> df -h <enter> gives:
> 
> /dev/hda1             6.7G  1.2G  5.1G  20% /
> NEW_DBS:SYS           8.0G  5.1G  2.9G  63% /.nwamd/NEW_DBS_SYS
> df: /mnt/server1: Input/output error
Yes, you get EIO on all operations to smbfs when the connection have been
lost and smbmount is gone.

> now in /etc/hosts I have:
> 
> xxx.yyy.zzz.2 inet_server_1 server1
> 
> and the command I used for smbmount is:
> 
> smbmount \\\\inet_server_1\\C$ /mnt/server1 -U <username>
But that is not 2.0.6 syntax ... You may have installed it, but you are
probably not using it. Also, consider mounting something other than c$,
you are not escaping the $ and I think you may have to do that. Add a new
share to the server for testing, if you have to.

(Why 2.0.6 and not 2.0.7? Shouldn't matter ...)

Also, you should be using the netbios name and not the DNS name. Is the
netbios name server1? then use //server1/share
(possibly with ip=xxx.yyy.zzz.2 if it can't find it from the name)
> now, the only difference is this server is not plugged into the same
> switch as inet_server_1 and the other three linux boxes...could this
> make a difference as to why this is happening?
No, I don't think so. You get browsing problems if they are on different
networks (unable to broadcast to each other), but different switches
should be fine.


/Urban

PS.
users@lists.calderasystems.com doesn't accept non-member submissions so
feel free to forward this there if you want to. I've dropped it from the
Cc.