We are having trouble restoring a backup from SFTP Server to Publisher

[astanislaus]

Please see attached PHOTO file.

=========================

CCMDB Restore status has been saying "ACTIVE" under the RESULT Column for about 12 hrs now. Still has not completed. The backup is running VERY VERY VERY slow.

Even the ones that say 100% and SUCCESS under the RESULT Column took MANY MANYHOURS.

There is no network problem because this lab test network just has one Cisco 2960 switch with one port going to Publisher and other port going to a SFTP Server running on a Laptop. We took the Publisher OFF it's production network to rebuild the server and restore the database and we are now stuck during restore.

The two ports on the switch have no input or output errors and are running at 100/FDX. Continuous pings of packet size 65000 bytes between this Publisher and the SFTP Server are running without any problem from both ends.

I suspect a H/W problem with this Server. But what H/W do I change? Any other suggestions.

Alphonse

[Rob Huffman]

Hi Alphonse,

Is there any chance that the Hostname has been changed. I have seen this

problem if the restore is not done to the exact Hostname as the backup

Also, could you provide the exact CUCM version?

Cheers!

Rob

[f.giordano]

Hi,

Already had one issue during a restore which stuck.

I opened a Case and the guy asked me to change my SFTP server.

I told him I used it for 6 years and never had any issue.

He told me to use FreeFTPd server... He was right.

Restore worked after that.

You can try.

Regards

Fabian

[cbonilla001]

I am having the exact same problem, where I'm trying to restore but it is getting jammed at 66 % see screen shot below.

I must say that it is a cluster of 1Pub and 6 SUB's  but I am not even able to restore the PUB. It has gone over 24 hours now and at the same spot. I am using the suggested FreeFTP server and I also tried others as well.

Any more ideas from you guys ? 

Regards,

Carlos

admin:utils disaster_recovery status restore

Tar Filename: 2013-02-26-01-00-12.tar

Storage Location: NETWORK

Operation: restore

Percentage Complete: 66

CCM   CALIDCUCCMPUB01   PLATFORM   SUCCESS   Mon Apr 15 11:44:15 EDT 2013   /var/log/active/platform/drf/log/2013-04-15-11-36-11_r_calidcuccmpub01_ccm_platform.log

CCM   CALIDCUCCMPUB01   CLM   SUCCESS   Mon Apr 15 11:44:19 EDT 2013   /var/log/active/platform/drf/log/2013-04-15-11-36-11_r_calidcuccmpub01_ccm_clm.log

CCM   CALIDCUCCMPUB01   TCT   SUCCESS   Mon Apr 15 11:44:20 EDT 2013   /var/log/active/platform/drf/log/2013-04-15-11-36-11_r_calidcuccmpub01_ccm_tct.log

CCM   CALIDCUCCMPUB01   SYSLOGAGT   SUCCESS   Mon Apr 15 11:44:20 EDT 2013   /var/log/active/platform/drf/log/2013-04-15-11-36-11_r_calidcuccmpub01_ccm_syslogagt.log

CCM   CALIDCUCCMPUB01   CDPAGT   SUCCESS   Mon Apr 15 11:45:12 EDT 2013   /var/log/active/platform/drf/log/2013-04-15-11-36-11_r_calidcuccmpub01_ccm_cdpagt.log

CCM   CALIDCUCCMPUB01   CCMPREFS   SUCCESS   Mon Apr 15 11:48:32 EDT 2013   /var/log/active/platform/drf/log/2013-04-15-11-36-11_r_calidcuccmpub01_ccm_ccmprefs.log

CCM   CALIDCUCCMPUB01   CCMDB   Active   Mon Apr 15 11:48:33 EDT 2013

CCM   CALIDCUCCMPUB01   MOH   ---   null

CCM   CALIDCUCCMPUB01   BAT   ---   null

[Aman Soi]

Hi,

What is the call manager version u are using as also asked by Mr. Rob?

Did the Publisher went down and that is the reason u are restoring BAckup?

If u have re-build the Pub, is it on same exact version anf credentials like OS/Application/Security password are same.

regds,

aman

[cbonilla001]

Hello Aman,

Let me give you more details. It is more of a trying out case

- I am migrating from MCS 6.1.5 to UCS 9.1.1a

- It is a phased migration so I have to have both environments running with intra-cluster, etc etc. That part is fine.

Here is the scoop. I have multiple offices with a very large number of Route Patterns, Translation Patterns. I noticed that 9.1.1 release provides you with more options to Export and then Re-Import to the system so I decided to try the following plan in order to save myself from having to manually enter thousands of patterns.

In an isolated Server/Environment

- Install same release of Call Manager 6.1.5xxxx (VMWare-UCS)

- Restore from MCS Running version the full backup

- Upgrade to 8.x or basically the path to get me from 6.1.5 to 9.1.1

- Once upgrade completed to 9.1.1, Complete an Export of all information that I require from the list of options Release 9.1.1 gives (much more than 6.1.5)

- Import that file with the information I selected to the new UCS 9.1.1 environment thus avoiding manual entry

I have the isolated host configure exactly the same: release + patch, matching IP and Hostname, etc etc.

The restores starts up just fine and within minutes gets to 66% but then it is just stuck there.  Note that I am aware that is not the Cisco recommended or supported approach but I thought it was worth the try given the amount of manual work I would be saving

Thanks again

Carlos

[Leo Salcie Tejeda]

Hi,

I have a issue like this some month ago. At the end was caused by the TFTP server because it wasn't design to transfer a lot amount of MBytes.

Try again running freeFTPd.

Regards

Please remember to rate useful posts clicking on the stars below.
Favor calificar todos las respuestas útiles.
___________________________________________
LinkedIn Profile: do.linkedin.com/in/leosalcie

[Aman Soi]

Hi,

The screenshot attached shows TFTP as SUCCESS.I have faced the similar problem while taking backup and was getting stuck at TFTP because of large number of files.So, unnecessary file swere deleted and backup was successful.

regds,

aman

[cbonilla001]

Hello,

Thanks a lot for the response. I will try that.

However, the other problem I run into is that when I try to access the Administration or Servicevality I get a DB error but I am still able to login to the OS administration and Disaster REcovery Pages

BTW Anyway to kill the Restore process? I have to basically rollback on my VMWare via snapshots

[Aman Soi]

Hi Carlos,

What I have understood is that u have build a separate setup on VMWare for CUCM version 6.1.5 and trying to restore the backup taken on MCS  which is getting stuck at 66%?

Secondly, u have mentioned for Route Patterns and Xlation Patterns . Why export is required when u have already taken backup and trying to restore ?

regds,

aman

[cbonilla001]

Hello Aman,

The backup I have from MCS is a full/succesful backup.   Where it is failing at 66% is when I try to restore that MCS backup to the  VMWare CUCM.

The reason why  I want to export/import is so that I can go site by site as I have to complete the migration in phases. The restore I am trying todo is to the 6.1.5 with all information. Once I have that I will migrate that VMware to 9.1. Then I would do Export of the information I need only so I can import that into my Production 9.1 environment which happens to be on VmWare as well.

I hope I am able to clarify my own question on that.

Thanks again and Best Regards

Carlos