09-26-2011 04:46 AM - edited 03-19-2019 03:40 AM
I am attempting to upgrade our unity connection boxes from 7.1.5 to 8.5.1.
I have legacy MCS 7845H2 - It has 4 disks - 2 mirrors of 72gb disks. 4mb of RAM and 2 X 2.3ghz processors
I have UCS C210 VMWare Hosts.
I have attempted 2 migration methods.
The first method involves - upgrading the existing system and restoring a backup of this to a VM
This fails with a hardware compatablity check.
The second method involves - backup existing systen - build VM at same level (using recommending OVA) - restore data onto VM - upgrade VM
This fails with a restore problem - I have used 'TITAN' ftp, but please note the backup is < 900mb.
The error is
Physical restore failed - could not fork server connection, SQLCODE -25596
Program over.
Restore failed during Unity Connection DB ontape restore operation.
----> BEGIN Standard Error
/opt/cisco/connection/lib/disaster-recovery/database_do_restore.real.py:32: RuntimeWarning: Python C API version mismatch for module iproduct: This Python has API version 1013, module iproduct has version 1012.
import iproduct
gunzip: stdout: Broken pipe
----> END Standard Error
Has anyone out there attempted to upgrade to 8.5.1 - please let me know how you got on.
Any help or ideas appreciated.
Solved! Go to Solution.
09-29-2011 10:52 AM
It's probably time to open a TAC SR on this if you haven't already. If the checksum validation passes and it won't restore the backup it's beyond what we can do on the forums.
09-26-2011 07:32 AM
Upgrading on the MCS hardware will fail because the hard drive mirror is not 146GB which is the minimum in 8.x.
Installing 7.1(5) on the UCS isn't supported by Cisco which means you're in uncharted/tested territory. Did you install exactly the same build (e.g. 7.1.5.10000-10) on UCS? If yes, have the services started, namely Informix?
You may have to take a backup on your MCS server, rebuild it on MCS with a 146GB RAID1 array, upgrade to 8.5(1), backup again, and finally restore on UCS.
09-28-2011 04:38 AM
You may have to take a backup on your MCS server, rebuild it on MCS with a 146GB RAID1 array, upgrade to 8.5(1), backup again, and finally restore on UCS.
Flattened my subscriber - reconstructed the raid to be 2 X 146gb
I built it as the publisher
7.1.5 installed fine
Installed language pack
I double checked all versions.
I attempted my restore - it failed again ...
so I didnt get to upgrade the box to 8.5.1 and take a backup !
09-28-2011 10:09 AM
What does the DRS restore log say?
09-29-2011 01:05 AM
Here is the end of the restore log ...
CCM DB is in On-Line status after 6 second.
Restoring the Unity Connection database from ontape archive...
Archive Tape Information
Tape type: Archive Backup Tape
Online version: IBM Informix Dynamic Server Version 10.00.UC9XF
Archive date: Mon Sep 26 19:45:06 2011
User id: root
Terminal id: ?
Archive level: 0
Tape device: STDIO
Tape blocksize (in k): 32
Tape size (in k): 0
Tape number in series: 1
Spaces to restore:1 [rootdbs ]
2 [cuc_er_dbspace ]
3 [dir16 ]
4 [rpt ]
5 [mbx16 ]
6 [dyn ]
7 [mbx ]
8 [dir ]
9 [log ]
10 [cuc_er_sbspace ]
11 [ciscounity_sbspace ]
12 [ciscounity_sbspacetemp ]
Archive Information
IBM Informix Dynamic Server Copyright IBM Corporation 1986, 2008
Initialization Time 06/01/2011 15:51:53
System Page Size 2048
Version 14
Archive CheckPoint Time 09/26/2011 19:45:07
Dbspaces
number flags fchunk nchunks flags owner name
1 40001 1 1 N B informix rootdbs
2 48001 2 1 N SB informix cuc_er_sbspace
3 40001 3 1 N B informix cuc_er_dbspace
4 40001 4 1 N B informix dir16
5 40001 5 1 N B informix rpt
6 42001 6 1 N TB informix temp
7 40001 7 1 N B informix mbx16
8 42001 8 1 N TB informix temp2
9 40001 9 1 N B informix dyn
10 40001 10 1 N B informix mbx
11 40001 11 1 N B informix dir
12 40001 12 1 N B informix log
13 48001 13 1 N SB informix ciscounity_sbspace
14 48001 14 1 N SB informix ciscounity_sbspacetemp
Chunks
chk/dbs offset size free bpages flags pathname
1 1 0 250000 183419 PO-B /var/opt/cisco/connection/db/root_dbspace
2 2 0 915000 176607 POSB /var/opt/cisco/connection/db/cuc_er_sbspace
3 3 0 85000 83217 PO-B /var/opt/cisco/connection/db/cuc_er_dbspace
4 4 0 128000 13824 PO-B /var/opt/cisco/connection/db/dir16_dbs
5 5 0 1024000 744409 PO-B /var/opt/cisco/connection/db/rpt_dbs
6 6 0 512000 511922 PO-B /var/opt/cisco/connection/db/temp_dbs
7 7 0 512000 495816 PO-B /usr/local/cm/db/informix/data/mbx16_dbs
8 8 0 512000 511897 PO-B /var/opt/cisco/connection/db/temp2_dbs
9 9 0 128000 110000 PO-B /var/opt/cisco/connection/db/dyn_dbs
10 10 0 128000 119774 PO-B /usr/local/cm/db/informix/data/mbx_dbs
11 11 0 1024000 362797 PO-B /var/opt/cisco/connection/db/dir_dbs
12 12 0 150000 24947 PO-B /var/opt/cisco/connection/db/log_dbs
13 13 0 10000 518 POSB /var/opt/cisco/connection/db/ciscounity_sbspace
14 14 0 450000 22554 POSB /var/opt/cisco/connection/db/ciscounity_sbspacetemp
Physical restore failed - could not fork server connection, SQLCODE -25596
Program over.
Restore failed during Unity Connection DB ontape restore operation.
----> BEGIN Standard Output
----> END Standard Output
----> BEGIN Standard Error
/opt/cisco/connection/lib/disaster-recovery/database_do_restore.real.py:32: RuntimeWarning: Python C API version mismatch for module iproduct: This Python has API version 1013, module iproduct has version 1012.
import iproduct
gunzip: stdout: Broken pipe
----> END Standard Error
09-29-2011 10:52 AM
It's probably time to open a TAC SR on this if you haven't already. If the checksum validation passes and it won't restore the backup it's beyond what we can do on the forums.
09-30-2011 01:17 AM
Thanks for the help - I am opening a case now.
10-12-2011 06:55 PM
Hi, Were you able to get some resolution on this. I am getting the exact same error for moving the DB for version 7.1.3 from one server to another.
Thanks,
Toor
10-12-2011 08:43 PM
Guys..I was running with this issue for last one week and i finally found a fix ..not sure if this will work for everyone
I was trying to upgrade 2.x via 7.1.5 to 8.5 unity connection on vmware desktop. I couldnt go beyond 7.1.5.
7.1.5 to 8.5 upgrade used to run for 35 min and then fail. We tried doing it on Cisco MCS 7825/7835 and even UCS C210 but nothing worked. Finally i read this document and wondered that its talkinng about more RAM and HD for CUC 8.6 so innstead of using 8.5 CUC/CUCM template i tried 8.6 OVA CUC template ( this creates a VM with 4 GB RAM and 160 GB RAID 1 HDisks and this worked.
One thing i am not sure is whether this will work on a vmware desktop if you increase RAM and HDs as specified in CUC ova 8.6 ...this is something to be tested.
http://www.cisco.com/en/US/docs/voice_ip_comm/connection/8x/upgrade/guide/8xcucrug010.html
Hope this will help !
10-14-2011 03:41 AM
Still no resolution on this issue.
I have a 7.1.5.300000-117 + GB Locale.
The backups work everytime.
I have tried restoring on multiple variants of hardware profiles - including the 8.6 OVA template. I have also tried using 8 X 72gb disks in a MCS 7845H2 server.
Also the same 'sql fork' error on restore.
10-23-2012 01:28 PM
I'm so glad newer versions of the UC product suite aren't as dependant on NTP and DNS reachability when installing/restoring/upgrading. That aspect of these apps was so planned by someone who is not in the field. It is totally inefficient...
10-17-2011 01:39 AM
I got a resolution to this issue - the DNS servers MUST be available and reachable for the restore to work.
It was as simple as that.
07-12-2012 03:39 PM
This is also relevant with a restore operation failing. By deleting dns, I was able to restore an 8.6.2 server in a lab environment. I should have figured it was that simple. Thanks Keanej! +5
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