OES2sp2 miggui Failed to mount volume

Question

Trying to migrate NW65SP8 (DS8.8.4) to OES2sp2.
miggui is running on the OES2sp2 server which is patched up to date using rug.
Source and target servers are in the came container in the DS.
I can define the source and target servers and get to the point where I want to configure the migration of volumes, but then I get a failure. The log (minus the timestamps for clarity) reads:
ERROR - FILESYSTEM:volmount.rb:Command to mount source: ruby /opt/novell/migration/sbin/volmount.rb -s 138.37.100.118 -a "cn=joshua,ou=sys,o=qmw" -c cp437 -f "/var/opt/novell/migration/cc18/fs/mnt/source" -m -t NW65
INFO - FILESYSTEM:volmount.rb:*****************Command output start**********************************
INFO - FILESYSTEM:volmount.rb:
INFO - FILESYSTEM:volmount.rb:Information: ncpmount using code page 437
INFO - FILESYSTEM:volmount.rb:Information: ncpshell command executed as: LC_ALL=en_US.UTF-8 /opt/novell/ncpserv/sbin/ncpshell --volumes --ip=138.37.100.118 --u="joshua.sys.qmw"
INFO - FILESYSTEM:volmount.rb:Information: Mounting Volume = _ADMIN
INFO - FILESYSTEM:volmount.rb:Information: Executing command: /opt/novell/ncl/bin/nwmap -s 138.37.100.118 -d /var/opt/novell/migration/cc18/fs/mnt/source/_ADMIN -v _ADMIN
INFO - FILESYSTEM:volmount.rb:Fatal: Failed to mount volume _ADMIN for server 138.37.100.118
INFO - FILESYSTEM:volmount.rb:Fatal: SystemCallError, Unknown error 1008 - Failed to mount volume _ADMIN for server 138.37.100.118 .
I tried executing what appears to be the offending command by hand:
ruby /opt/novell/migration/sbin/volmount.rb -s 138.37.100.118 -a "cn=joshua,ou=sys,o=qmw" -c cp437 -f "K" -m -t NW65 -p password --debug
which produces the output:
Information: ncpmount using code page 437
Information: ncpshell command executed as: LC_ALL=en_US.UTF-8 /opt/novell/ncpserv/sbin/ncpshell --volumes --ip=138.37.100.118 --u="joshua.sys.qmw"
Information: Mounting Volume = _ADMIN
Information: Executing command: /opt/novell/ncl/bin/nwmap -s 138.37.100.118 -d K/_ADMIN -v _ADMIN
Fatal: Failed to mount volume _ADMIN for server 138.37.100.118
Fatal: SystemCallError, Unknown error 1008 - Failed to mount volume _ADMIN for server 138.37.100.118 .
Information: File K/_ADMIN/Novell/Cluster/PoolConfig.xml does not exist. No cluster resources attached
Information: Mounting Volume = SYS
Information: Executing command: /opt/novell/ncl/bin/nwmap -s 138.37.100.118 -d K/SYS -v SYS
Fatal: Failed to mount volume SYS for server 138.37.100.118
Information: unmounting all mounted volumes
Information: Executing command:/opt/novell/ncl/bin/nwlogout -f -s QMWCC18
Cannot perform logout: Cannot connect to server:[QMWCC18]. Error:NWCCOpenConnByName:
Information: Command Output:
Information: Executing command: rm K/*
rm: cannot remove `K/*': No such file or directory
Fatal: SystemCallError, Unknown error 1008 - Failed to mount volume SYS for server 138.37.100.118 .
SLP shows both servers, they are on the same network as each other, and the firewall is turned off.
Does anyone have any idea what may be causing the mount failure or what error 1008 might be?
Tim

Answer 1

cgaa183 wrote:
> Thankyou all for helpful suggestions, I'll go through them.
>
> 1. novfsd
> /etc/rc.d/novfsd status
>
> running
>
> /etc/rc.d/novfsd restart
> Stopping Novell novfs daemon...
>
> done
> Starting Novell novfs daemon...
>
> No Config File Found - Using Defaults
> novfsd: Novell Client for Linux Daemon
> Copyright 1992-2005, by Novell, Inc. All rights reserved.
> Version 3.0.1-503
>
> done
>
> Nothing changes, so I don't think that was the problem in this case.
>
did you restart the server after installing OES2 SP2?
> 2. try using migfiles
> /opt/novell/migration/sbin/migfiles -s 138.37.100.118 -v APPS -V APPS
>
> The result:
> Error:
> Error: nbackup: Unable to retrieve the Target Service Name list from
> 138.37.100.118
> Error:
> Fatal: nbackup command failed to execute: nbackup: Connection denied
>
migfiles is not able to connect to the TSA on the source server. Either
TSAFS is not loaded on the source server or not able to locate the TSA
using SLP.
> This might be informative to someone who knows a little more than me. I
> wonder if I can call it with options to get more information. I have
> just re-checked I can attach and mount volumes with the same username
> from another system.
>
> 3. ruby version
>
> rpm -qa |grep ruby
> ruby-1.8.4-17.20
> rubygems-0.9.2-4.4
> rubygem-needle-1.3.0-1.5
> rubygem-net-sftp-1.1.0-1.5
> ruby-devel-1.8.4-17.20
> rubygem-net-ssh-1.0.9-1.5
>
> Doesn't appear to be an afflicted version and migfiles --help tells
> me all about the options available.
>
>
> As #2 looked interesting I thought I'd look at it a bit more. I turned
> up TID 7001767. The reason migfiles failed for me was that SMDR and
> TSAFS weren't loaded, at least I don't get the error and file migration
> appears to start now they are loaded, though it does appear to have
> ceased copying files rather prematurely....
>
> Going back to volmount.rb I now realise its using ncpfs, and not a
> lightweight Novell client. So I tried mounting a volume by hand:
>
> qmwcc28:~ # ncpmount -S 138.37.100.118 -U joshua.sys.qmw K
> Logging into 138.37.100.118 as JOSHUA.SYS.QMW
> Password:
> ncpmount: Server not found (0x8847) when trying to find 138.37.100.118
>
> A bit of a giveaway, but why doesn't it work?
> It seems I need to use -A DNSname -S servername and then it works.
> The next important bit seems to be
> /opt/novell/ncpserv/sbin/ncpshell --volumes --ip=138.37.100.118
> --u="joshua.sys.qmw"
> which executed by hand lists volumes correctly with the output:
> Please enter your password ?
> [0] SYS
> [1] _ADMIN
> [2] APPS
> 3 NCP Volumes Mounted
>
> "ncpshell" appears to be from Novell's client for Linux so I don't
> understand why we'd be trying to use that if we're using ncpfs, and we
> already know which volumes are mounted by looking in the folder in which
> we mounted the server using ncpfs. AFAICS is used to invoke NLMs on
> NetWare remotely using OES so its not testing anything we don't already
> know.
>
> This takes me inevitably to "nwmap". "nwmap" is also from Novell's
> client for Linux so maybe the ncpfs stuff is unnecessary.
> /opt/novell/ncl/bin/nwmap -s 138.37.100.118 -d sys -v SYS
> produces:
> map: server not Found:138.37.100.118 - drive sys not mapped
>
ncpmount uses udp as the default. Add the option -o tcp to the ncpmount
command then mount should work.
> nwmap doesn't ask for a username. Maybe I'm wrong, but as far as the
> Novell client goes I don't think it can have attached or logged into the
> source server (ncpfs having a different connection table and ncpshell
> asking the remote server to return the answer). I can't actually see
> where /volmount.rb is calling nwmap at the moment but the results I get
> my calling it at the command prompt with the same options given in the
> log are the same.
>
if there is an existing connection to the same tree, nwmap does not ask
for user name. Use the command "nwconnections" to check the existing
connections. Use nwlogout to logout the connection. check
/var/opt/novell/nclmnt/root/ for any stale entries.
> I've tried logging in using nwlogin, but that fails too saying:
> Cannot perform login: The system could not log you into the network.
> Make sure that the name and connection information are correct, then
> type your password again.
>
> ncl_scanner -T does list NDS trees but I suspect its only querying an
> SLP server and nothing more useful. ncl_scanner -S produces:
> INFORMATION FOR SERVER [QMWCC18] :
> Server Name : [QMWCC18]
> Bindery : [FALSE]
> eDirectory Context : []
> should it show a context?
>
> Looking at the files of the Novell client on the system, it looks a
> rather cut down set with no config files. Even having introduced
> protocol.conf the situation is not improved, but I'm now sure the
> problem lies in this area. Possibly a full client installation is
> required, or maybe there is something else wrong which is preventing the
> client from working correctly. namcd is looking suspect.
>
>
You do not need all files for Novel Client. If you want
You can logout all connections using the command "nwlogout -a" and try
the nwmap command again.
"/opt/novell/ncl/bin/nwmap -s 138.37.100.118 -d
/var/opt/novell/migration/cc18/fs/mnt/source/SYS -v SYS"
Looks like the novell client is failed to resolve the IP address. You
can do the following to configure different name resolution methods in
the following way:
Create a file: /etc/opt/novell/ncl/protocol.conf with data.
Name_Resolution_Providers=NCP,SLP,DNS
then restart rcnovfsd deamon using the command "rcnovfsd restart"
do you see any NCP errors in the network packet trace?
regards
Praveen