11-27-2007 01:13 AM
Hi all
After upgrading RME to 4.0.6 nothing works for SSH devices Archive,Netconfig). I get "RME_CDL1031:Transport session to device failed Could not detect protocols running on the device TELNET: Failed to establish TELNET connection to 10.69.xx.x ".
This was working fine before upgrade from 4.0.5, also the credentials report shows "Failed to connect". I know the credentials are ok since i enabled Telnet in one of the routers to prove credentials OK.
I also tried on a testserver, everything is fine even after CS upgrade to 3.0.6 but after applying RME 4.0.6 I get the same symptom.
Using sniffer i can se the routers responding with "SSH version 1.5" but then it fails.
anyone got a brilliant idea ?
rgds
Mats
Solved! Go to Solution.
12-04-2007 07:37 AM
My patch is available, and does work. If all your devices fail after applying it, the patch must have been corrupted in transit, or was installed incorrectly. That said, the RME 4.0.6 update was just re-posted with my patch. So if you re-download and install RME 4.0.6, the problem will be fixed.
11-27-2007 01:26 AM
Hi Mats,
Enable SSH in RME -> Admin -> configMgmt ->Transport settings -> Archive mgmt/netconfig.
HTH
sathappan.s
11-27-2007 01:33 AM
Hi
Thanks, but as I said it worked fine in 4.0.5 and I checked the Transport settings twice. I can tell by sniffer that RME (or credential verification) is trying SSH but then fails.
rgds
Mats
11-27-2007 09:38 AM
This error indicates a problem with the data returned from the TCP connection to port 22. You should enable ArchiveMgmt Service debugging under RME > Admin > System Preferences > Loglevel Settings, run a new archive sync operation, then check the dcmaservice.log for errors.
Of course, I'm assuming that before you upgraded to RME 4.0.6, you applied the Common Services 3.0.6 update. If not, you will need to apply the CS 3.0.6 update, then re-apply the RME 4.0.6 update.
11-27-2007 02:28 PM
FYI, I tested this locally after upgrading to CS 3.0.6 and RME 4.0.6, and I am not seeing any problems with SSH.
11-27-2007 11:37 PM
Hi
Thanks for responding quickly.
Yes i have confirmed CS 3.0.6 is ok installed.
Attached output in dcmaservice log after enabling debug. From what I can tell it first says "SSH is running on /10.69.77.1" and then shortly after a FATAL "could not detect protocols running on the device". I have to admit it is not my best skill to read theese debugs, please have a look to see if you can pinpoint the problem.
Btw
I reinstalled RME 4.0.5 on the testserver and synched config on the same 10.69.77.1 router without problem.
rgds
Mats
11-28-2007 12:00 AM
Looks like your installation is bad, or you have re-applied some old patches after installing RME 4.0.6. The NMSROOT/MDC/tomcat/webapps/rme/WEB-INF/classes/com/cisco/nm/rmeng/util/rmedaa/RMEDeviceContext.class looks bad. The line numbers in your log do not line up with the actual code, but they do on my server.
You might try reinstalling the RME 4.0.6 update. The MD5 checksum for the file above should be:
4d04e1ad1e29a206899c0853857b2fb4 RMEDeviceContext.class
11-28-2007 04:43 AM
Hi
I reinstalled 4.0.6 on the testserver (Windows 2003, as well as Prod server) no additional patches on the testserver only CS and RME. sych failed again.
The RMEDeviceContext.class got a different checksum on my system, but is the same on both servers as well as timestamp and bytesize.
But the culprit must be RMEDeviceContext.class, as I did an "ugly" test on the testserver and swaped it to 4.0.5 version and restarted, archivesynch is ok after that.
Fileinfo:
ab9c8b46e27a87719d01f66b8343f9da 2007-11-15 11:33 22Â 009 RMEDeviceContext.class
any idea ?
rgds
mats
11-28-2007 10:09 AM
I have a strong feeling the Windows RME 4.0.6 is broken. I assume you're testing on Windows?
11-28-2007 10:39 AM
Yes, and the "live one" is w2003 also.
It seems like using version 4.0.5 file is a temporary workaround to this problem. I have not noticed any other probs with RME yet, been focusing on this...
pls let me know what happens.
brgds
mats
11-28-2007 10:55 AM
I have confirmed the Windows update is bad. If you download the Solaris update, extract it, you will find an RMEDeviceContext.class. That is a good version. Simply replace your version with that one, restart dmgtd, and you'll be fine.
11-28-2007 11:42 PM
Hi
I get the same error with the solaris one on both servers, I verified checksum on the solaris file and that is the same as yours.
I suppose i can use v 4.0.5 until a fix is available, to avoid reinstalling RME on the Prod server.
rgds
mats
11-29-2007 08:35 AM
I see the problem. Please open a TAC service request, and have your engineer contact me for the patch.
11-29-2007 11:13 PM
Hi
I will do that. Thanks a lot for pointing out the problem so fast.
brgds
Mats
12-04-2007 06:43 AM
Was a patch made available for this problem or was a bugID assigned to it?
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide