Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 
Users might experience few discrepancies in Search results. We are working on this on our side. We apologize for the inconvenience it may have caused.
New Member

Unity Connection failure

Hi All, All this migration stuff is driving me nuts. Latest failure is connection. Managed to upgrade publisher to V10.0 from 8.6 and everything seemed fine, started on the subscriber and it failed which reverted back to 8.6 so looking at the publisher, I now have an error message that says "Unity Connection cannot accept new messages and you may not be bale to upgrade connection" This wasn't  there after the PUB upgrade and it all seemed to work fine - any ideas folks? 

I highly suspect that i'll have to use COBRAS to export and do a fresh install unless anyone has a bright idea.






Hi Paul,This error message

Hi Paul,

This error message could be due to high disk usage on the Publisher. Can you check the output of 'show status' from the Publisher. Also, did you collect the install logs for the subscriber install failure to check what may have caused the issue.



New Member

Hi Folks, well just to answer

Hi Folks, well just to answer, it seems yet again that there are serious bugs in migrating from 8.X to 10.X. I ended up using COBRAS to back up then do a fresh installation of 10.X. Even TAC couldn't fix it so after 3 weeks of trying, I did what i wanted to do after 3 days and it works.


Lesson learnt - always do a COBRAS backup.

To clarify, are you seeing

To clarify, are you seeing that error on your Publisher or on your Subscriber?
How long was the gap between upgrading the pub and attempting the sub? Is your freshly upgraded server properly licensed (i.e. you did the conversion)?
As Manish mentioned, you should check your disk as well to see if it's full. SSH or open up the console to the server and run 'show diskusage common'. you will get an alert about the command taking a long time, press 'y' to accept and look at the top of the output to see the percentage of utilization.

What size deployment is this for, or which OVA is being used? It looks like the resources shouldn't need to be increased for the VM going from 8.6 to 10.0. Unless you were getting lucky on 8.6 running off an OVA for 8.5 that required less users (i.e. 5000 users on 8.5 required 4GB memory and on 10 requires 6GB).

Tons of little things that could pop out. I see it's been a week since you've posted though. Reply and let us know how it's going.




CreatePlease to create content