CUCM 9.1.1.20000-5 to CUCM 10.5.1.10000-7 Upgrade Failure

[Conwell Less]

Hi,

I am performing an upgrade on a UCS Server from CUCM 9.1.1.20000-5 to CUCM 10.5.1.10000-7 and CUCM10.5.1.11900-13. Each time i get the error: "The selected ISO file is not valid. This indicates that the download of the ISO file corrupted the file or the ISO file on the remote server is corrupt."

I have tried installing the file ciscocm.version3-keys.cop.sgn first as well i get same error. I have downloaded this software twice from Cisco downloads directly and via PUT.

Screenshot is attached with error. Any help will be appreciated.

Regards

[Aman Soi]

Hi,

I was going through the discussion but u have already tried uploading ciscocm.version3-keys.cop.sgn

https://supportforums.cisco.com/discussion/12243401/md5-error-when-upgrading-105

Did u verify the MD5 checksum of files/patches before uploading?

regds,

aman

[Conwell Less]

yes, please see response to Ayodeji Okanlawon

[Conwell Less]

hi,

sorry for only responding now, there was a problem with my hardware. i successfully manage to install 10.5.

thanks for all your help.

[Aman Soi]

thanks for update.

If u could share , what exact was the issue in Hardware would be helpful in future.

regds,

aman

[Conwell Less]

my lab test was on unsupported cisco hardware. i had a dell server i was testing on. then rebuild my lab on the ucs and had no problem.

[Ayodeji Okanlawon]

Two things to do

1. Check the checksum file and ensure it is the same as the one on cco

2. I see the ff: errors on your logs. This suggest that you do not have enough space in the common partition for this upgrade. You need a minimum of 29GB in the common partition for a successful upgrade. CUCM needs to copy the 5GB file first to this partition. Use RTMT to determine the space you have in the common partition and delete log files to free more space.

10/22/2014 07:52:50 upgrade_validate_file.sh|(CAPTURE) Free space: 51316539392, Needed space: 1992138752|<LVL::Debug>
10/22/2014 07:52:50 upgrade_validate_file.sh|There is enough space on device to proceed.|<LVL::Info>

Please refer to this thread on best practices on upgrades..

https://supportforums.cisco.com/discussion/12313756/cucm-862-upgrade-cucm-105

[Conwell Less]

Hi Guys,

So I have been doing lots of reading and investigating before I decided to reply.

What I find is the following:

1. The MD5 Checksum on the Cisco CCO and on my PC is exactly the same and I have checked this using a MD5 tool. However, when I run the upgrade, the MD5 seems to be different (Do not see error about that).

2. I also think I may have insufficient space available and have verified this using RTMT. I tried to locate the LogPartitionLowWaterMarkexceeded and LogPartitionHighWaterMarkexceeded alerts in RTMT to adjust the thresholds but I do not see these alerts either in the logs.

3. I have adjusted the 9.1 ova template to match 10.5 requirements as discussed in the Cisco document and still get same error. Changed OS, VM Nic adapters, etc.

4. I have loaded the ciscocm.vmware-disk-size-reallocation-1.0.cop.sgn file as well.

Everything I am referring is in the doc attached. I am out of ideas now.

[Ayodeji Okanlawon]

First of all fix your disk space issues. Delete log files and purge CDR if you have any

1. Delete all log files on all cucm servers and purge CDR records

   file list activelog /cm/trace/ccm/sdl/*

   file delete activelog /cm/trace/ccm/sdl/* noconfirm

2. The cop file is used to increase disk size not to free space, so that wont help

3. To change the logpartitonhighwatermark threshold go to alert central>right clock on the logpartitonhighwatermark alert and click on set properties. Select the servers you want to change this for, click next and use the slider to change the threshold. Do note the value before chaning it, cos you need to change it back after upgrade. Do the same for the lowlevel alert

4. Try the upgrade again

5. after deleting the log files use RTMT to check that you have 29 GB free in common partition

[Conwell Less]

just also for information purposes, this is a clean install of 9.1 which has nothing on it. I had this error when trying to do the upgrade on a live system when i got this error.

Then installed the same instance on a VM in my LAB to simulate (the lab cucm has nothing configured on it. As in my previous post, those alerts are not seen in alert central and cannot change the properties if its not there.

Will still try and delete the log files anyway

[Ayodeji Okanlawon]

In that case I suggest you go to cco and download the cucm iso from there. This may take a while depending on your internet speed. If you do not see the alert, then something is not quite right with your cucm install..

[Conwell Less]

that is why i am so confused, i have downloaded this twice already. It works quite fine when i do a fresh install but does not work with an upgrade.

[Ayodeji Okanlawon]

That is a little strange..Do you have webex? Can you send me a webex invite?

[Aman Soi]

Just to add: Are u trying to upgrade using the patch which u have recieved after u placed the order through UCSS.Have u opted for Physical Delivery?

regds,

aman

[Ayodeji Okanlawon]

Conwell,

Have you managed to resolve this? Can you please update us on your progress