Highlighted
Absent Member.
Absent Member.
3183 views

Cifs Problem on cluster

Hello All,
I am completly stuck...been on the forums and cannot figure out what is wrong.
From the start. We had a netware 6.5 so8 cluster running no problem. Just NSS no CIFS. New boss comes in and wants to move to AD. I recommend keeping the clusters and upgrade to Suse Linux 10 sp3 (so much better with our IDM solution as well) and putting CIFS shares on the servers so that MS clients can access their drives.
Well upgraded to SuseLinux SP3 OES2 now SP3. When i add the cifs protocol and restart the cluster resource goes to comotose. Thninking this is the same error as previously documented have tried running cifspool.py :
python cifsPool.py cn=ZENVOLUME_SERVER,cn=STCL,ou=STU,o=CAMPUS ZENVOLCIFS ldaps://127.0.0.1:636 cn=admin,o=campus PASSWORD

comes up with error :

INVALID_DN_SYNTAX: {'info': 'NDS error: illegal ds name (-610)', 'desc': 'Invalid DN syntax'}

the cluster object is correct... what am i doing wrong.... i can add the log files if needed

Thanks all...
Labels (1)
0 Likes
13 Replies
Highlighted
Micro Focus Frequent Contributor
Micro Focus Frequent Contributor

Try the modified script in TID 7005192. If the pool was created with NW
or OES2SP1 or earlier, this should work.
0 Likes
Highlighted
Absent Member.
Absent Member.

Thats the one i ran... and getting the error with that is shown before

Thanks
0 Likes
Highlighted
Absent Member.
Absent Member.

cn=ZENVOLUME_SERVER,cn=STCL,ou=STU,o=CAMPUS

This is indeed invalid, you can't have a CN inside a CN.

Andrew C Taubman (Sorry, support is not provided via e-mail) Opinions expressed above are not necessarily those of Micro Focus.
0 Likes
Highlighted
Absent Member.
Absent Member.

says in the documentation that the intial object is a cn AND the cluster object is a CN ... i couldn't figure that one out....
0 Likes
Highlighted
Micro Focus Frequent Contributor
Micro Focus Frequent Contributor

Actually, you can have objects inside a CN. The cluster object is a CN
and it has objects under it, such as the cluster resource and node objects.
0 Likes
Highlighted
Micro Focus Frequent Contributor
Micro Focus Frequent Contributor

Just to be clear, you copied the contents of the Additional Information
from the TID to a file on the server, made it executable, and ran it
with the specified parameters? If so, please do the following:

Please post the load script for the resource. Make sure CIFS is enabled
still before getting the script.

What nfap attributes are on the resource's virtual server object with
what values? You should have nfapCIFSAttach, nfapCIFSComment and
nfapCIFSServerName

Please post the entry from
/var/opt/novell/log/ncs/ZENVOLUME_SERVER.load.out from it going comatose.
0 Likes
Highlighted
Cadet 1st Class
Cadet 1st Class

Hello,

Were you able to solve this problem? I have the exact same problem and my CIFS Virtual Server name is not visible through the windows explorer and I cannot map to the virtual server name, either with "net use x: \\servername\share" or explorer. If I use "net use x: \\<IP-Address>\share" it works fines...???

As stated before I've tried both the original cifsPool.py and the "updated" cifsPool.py both give me the same error. I'm also sure that the context I've given in the command is correct. There is only one context for eDir objects. I've tried it with commas and periods, no luck...

Any help would be greatly appreciated.

A side note, my cluster resources do not go comatose after running the skript(s). They did before, so it seems like the skript fixes the "comatose" problem but it doesn't seem to fix my "Virtual server name Bind problem". When I look at the Virtual Cluster resource server object in eDir all the needed attributes are there and correct...???

Thanks in advance...
Mike
0 Likes
Highlighted
Cadet 1st Class
Cadet 1st Class

Oh btw, we have OES2 sp3 and the April Maintenance patches have been installed ("eDirectory 8.8 SP6 Patch2 for OES2 SP3 20110422" & "April 2011 scheduled maintenance for OES SP3 20110421").

Thanks...
0 Likes
Highlighted
Knowledge Partner Knowledge Partner
Knowledge Partner

Do you have DNS entries created for your CIFS server names? By default it'll add a "-w" at the end.

If they're not in DNS, then if you have WINS, you'll need to put the WINS server IP in the physical cluster node server object (NOT the virtual server object) and that should kick in

If you can map via the clustered IP then you know it's a WINS/DNS issue.
0 Likes
Highlighted
Cadet 1st Class
Cadet 1st Class

Hi,

Thanks for your prompt reply. I'll give that a shot (DNS/WINS). At the moment I'm not at the customer so it'll have to wait until tomorrow.

I also spoke too soon when it comes to the scripts solving the "Comatose" cluster resource problem. It appears as if the scripts do not solve it. 😞 I setup cifs on a cluster resource that I could regularly restart and after several days and a multitude of attempts it finally could be migrated or taken offline and brought back online without going comatose. The problem is I'm not sure exactly how I got it to work and believe that I just got lucky. I figure the following.... In iMgr go to the Cluster Manager and open up the properties of the cluster resource, there activate CIFS (under Protocols), then take the cluster resource offline, run the cifsPool.py script (brings INVALID_DN_SYNTAX Error, although the syntax IS correct) then check in iMgr / C1 the OTHER Properties of the cluster virtual server object and bring the cluster resource online. Unfortunately this didn't work.... Comatose. Therefore I used the "updated" cifsPool.py script and got the exact same error and the exact same problem. Then I tried several manaul attempts at it (loaded the resource w/o CIFS Protocol and performed the novcifs command per hand... seemed to have worked..... although if I activate CIFS again in the properties of the cluster resource in iMgr again, it goes comatose... again..... WTF???? You've got to be kidding me?

Question, is there a way to add the CIFS protocol manually and avoid using iMgr? If so, how? I'm assuming that I could edit the actual cluster load scripts but I have the problem that this is a production cluster and the customer wants to see it work in the manner that Novell's documentation says it'll work. Especially considering that they eventually will be trying to set several CIFS shares themselves.....

Thanks for any help and could you please inform me if Novell finally fixes this problem....

Mike
0 Likes
Highlighted
Knowledge Partner Knowledge Partner
Knowledge Partner

To be honest, I'm not sure what the cause/fix for this is yet. I just found a workaround. I do have an SR open but it's not going very fast (although in fairness I did tell the NTS person that it wasn't a "high" priority since I had a workaround). I'll see if I can get a bug/defect or something and if so, what a possible fix/patch may be.
0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.