Unsolved
This post is more than 5 years old
29 Posts
0
24787
June 12th, 2013 11:00
Cannot access CIFS shares after installing 2008R2 server.
I've deployed a new Windows 2008R2 server in an existing domain with 2 other non-R2 2008 DCs, but my users are having trouble accessing a the EMC fileshare when they are connected to this new server. They cannot access other non EMC shares just fine.
Does anyone have any ideas on what could be wrong?
Here are some details:
1. New server has roles of DNS, DHCP and AD and looks successful from DNS test and from replication of DHCP and AD from other servers.
2. Our celerra file share is named "fileserv" and when users tried to access it through existing drive mapping, they get prompted for Windows Credentials and says "Access is Denied".
3. If I try to access by running \\fileserv, they get the same error.
4. ping tests to fileserv and ip address are successful.
5. If I try to access by running \\ipaddress, they get the same error.
6. "fileserv" name in DNS actually points to same ip address as "emc-san1" (another alias)
7. If I try to access files by using \\emc-san1 , shares ARE accessible and normal.
8. If I try the "net use" cmd, I see that the existing drive mappings to "fileserv" say they are Unavailable.
8. We have load balancing dhcp servers and only users connected to this dhcp server (the new R2 server) are getting this error. Users still connected to the older non-r2 servers are connecting just fine.
9. I even just tried creating a new DNS record called "testserv" and tried to access file shares via \\testserv and get the same error.
10. Just ungraded the software on the celerra from 5.6.47 to 6 and I'm still getting this issue.
11. We have a printer here that has the Scan to File option where it will scan a document and upload it our fileserver while connecting to it via SMB. This feature no longer works after introducing this R2 server into our domain.
12. It's bugging me because ONLY access to the EMC is wonky. All non EMC shares work fine. If I reboot the R2 server, affected users will work for a little while, but then some users can't connect anymore. It almost seems like it is intermittent, or there is a connection limit and then blocks more users.
For the current workaround, I'm remapping to the "emc-san1" for my users to access files.
Note: I still want my users to ultimately connect to fileserv instead of emc-san1 because we might be removing the emc san soon and I can point "fileserv" to the new storage device.
Any ideas on this?
dynamox
9 Legend
•
20.4K Posts
0
June 12th, 2013 11:00
Look at SMB signing on the server, might need to disable it
JAssociates
29 Posts
0
June 12th, 2013 13:00
I'm running 6.0.70-4..just upgraded last week...per your recommendation too! Thanks Rainer
As for the server_cifs comment, do you know how do I check on that? This 6.x DART is much different from 5.6.47 that I was running last week..
Rainer_EMC
4 Operator
•
8.6K Posts
0
June 12th, 2013 13:00
What specific DART version are you using ?
What does server_cifs say in terms of config info for your CIFS server ?
Have you tried nas_checkup? It catches a lot of common config mistakes
bergec
275 Posts
0
June 13th, 2013 03:00
Also the server_cifssupport command as some options to "ping" Domain Controllers and check protocols
See results in server_log
Claude
Rainer_EMC
4 Operator
•
8.6K Posts
0
June 13th, 2013 03:00
The server_cifs command should be the same – its been around forever
See “man server_cifs”
If you don’t already have it I would suggest to download the CIFS manual as well as the technote for server_cifssupport
christopher_ime
2K Posts
1
July 15th, 2013 22:00
JAssociates,
Just curious were you able to resolve this? I suspect you either were asked maybe to rejoin the CIFS server (having since added a new DC), or instead prompted to run the following command:
.server_config -v “pdc dump”
Then if not listed, worked with support to maybe manually add the unlisted DC via a slight variation of the above command? Sorry, intentionally being elusive.
If so could you share your resolution? Another user I suspect is in the same situation but in a slight variation of the same (new DC introduced after CIFS server was joined).
Cannot access CIFS Shares after re-adding DC roles to Server 2008
Then again, I could be completely on the wrong track with this.
christopher_ime
2K Posts
0
July 15th, 2013 22:00
Lol, okay I just realized you were the author of both posts. Let's start with the output of the above command.
christopher_ime
2K Posts
0
July 15th, 2013 23:00
Agreed. I'm almost certain also that it has nothing to do with it being R2. I'm fairly certain it is related to introducing a new DC after the initial join of the CIFS Server. There is a command (slight variation of the harmless query above) to manually add it, but would like for support to verify first. Then again, as Rainer suggested unjoining and rejoining probably would resolve it also (if it is as I suspect), but of course would require downtime.
JAssociates
29 Posts
0
July 15th, 2013 23:00
Have you rejoined domains before?
Is it pretty straight forward with a command or two, or is it pretty complicated?
How much downtime is expected? I’m curious if I call EMC tomorrow, they’ll schedule me something later in the evening? If not, I’d maybe look at doing this myself. I need to resolve this and get on with my life!
JAssociates
29 Posts
0
July 15th, 2013 23:00
Hahahah Christopher..that’s funny! But you’re also awesome for doing all this research!
Some backstory: I originally installed a new 2008R2 server and starting getting these errors. So I backtracked and un-retired an older 2008R1 DC to restore functionality, but started to get these errors on that server, so I ruled out any possibility of a funky issue caused by R2.
I still haven’t found a resolution on these forums or from 2 support tickets with EMC.
I haven’t been suggested by support to rejoin the domain or run your command, but someone else mentioned it on the forum earlier today. Since I’ve never done that before, I’m going to call support to have them do it. Better be safe than sorry.
I’ll try your command in the morning and post my results.
Thanks for your persistence!
JAssociates
29 Posts
0
July 16th, 2013 15:00
Im having EMC support contact me tonight to do the re-joining to the domain. Hopefully, everything will be fixed after that. So I’ll post tmrw to let you know how it went…
JAssociates
29 Posts
0
July 17th, 2013 16:00
So EMC helped me out last night, but it didn’t seem to work.
They first stopped and restarted the CIFS server, but that didn’t do much, so they re-joined the domain. He noticed that after doing so, the log shows that all 3 servers were authenticating against AD. (They had the ‘>’ next to each name). However, after going back in the office today, we still had the same issue.
Do you have any other ideas on what to do? We’ve already upgraded the DART version and re-created the CIFS shares beforehand…
Thank you!
Archuperi721
110 Posts
0
July 25th, 2013 00:00
As you have mentioned the issue is intermittend, I would like to ask you whether you have any user limit value applied on the CIFS share. If then please clear / adjust the value accordingly. Have you ever applied any restriction on the file system, like IP range / ownership
JAssociates
29 Posts
0
July 25th, 2013 01:00
I was wondering the same thing, actually.
Would you be able to guide me and tell me how to check?
Specifically, I noticed that if I reboot the server, many users will be able to log on and access the files at first, but after a while, users will then get blocked again. If I reboot the DC, the same thing will happen each time..
kajibade
33 Posts
0
July 25th, 2013 04:00
Check the SMB version listed in the "Max protocol" field of the server_cifs command output. From what you describe, it sounds like the introduction of Windows 2008 R2 DC has pushed the default protocol disposition of the domain up to SMB2.
I wonder if the Celerra CIFS server configuraion is still running with NT1 protocol, inherited from earlier code releases.