I have 2 CM servers in a cluster both with 3.09 running on them. When I start up Web Attendant I get the User Directory but I can not log in to Web Attendant. If I point to the Primary CM I get no failure message when I try to log in. If I point to the Redundant CM I get and error message that says "A user with your ID is already logged in". Now if I stop the Call Manager services on one of the servers Web Attendant works just fine. I have heard there is a bug causing this but I can't find it anywhere on CCO. Any help would be appreciated.
A colleague of mine who is better with MS win 2k than I am found that one of the Data bases on the Primary CM was not replicating at all. We still have to figure out why. Here is how we did the work around.
1 create a new Call Manager group with the servers reversed (redundant then primary)
2 create a new Device Pool that uses the new Call Manager Group.
3 For the phones that will be using web attendant have them use the new Device Pool
4 Restart the phones that were changed in the prvious step.
5 In the settings of Web Attendant desktop client and the TSP point everything to the redundant Call Manager. Make sure the path for the user data base is also looking for it on the redundant Call Manager.
6 In Call Manager Service>Cisco Web Attendant>Pilot Point Configuration set the primary Call Manager to the IP address (or DNS name) of the redundant Call Manager.
7 On the Redundant Call Manager go to Service>Control Center and stop then restart the Cisco Telephoney Call Dispatcher.
8 Restart the Web Attendant on the users PC and login and then "go online"
This is what we had to do to get it to work. Screwy I know, but hey it is working for us. We are still looking into the data base issue when time permits. Hope this is helpful.
I noticed that in your step one that you are actually making the redudant CM the server that the phones register, thereby forcing it to the "primary" call manager. In my configuration, for example, CCM001 is the Publisher and Redundant call manager. CCM002 is the Subscriber and Primary call manager. The TFTP, DHCP and replication (so they say) is done from CCM001 and the call processing is done by CCM002. If I read your insruction correctly this would make my Publisher the Primary and the Subscriber the Redundant. I wanted to have one server dedicated for call processing and one for database replication.
I will give it a try. I also noticed that if I am in the Subscriber and search for users from the CCM admin page I get an error. This is probably due to the database replication issue. Have you had any problems with the softphone and dual call managers?
TAC has given me a procedure to fix the database issue, I have yet to try it though.
Are you having errors when you log into the subscriber (In my case the subscriber is the Primary CM and the Publisher is the redundant CM) and select user or global directory? How about if you are to log into DC directory Administrator from the subscriber server and enter the default profile then logging in as "Directory Manager" password "cisocisco"? I was having all of these problems, including web attendant.
Although your solution worked I was still having problems with the database integrity on the Subscriber. The user info was not replicated from the publisher. Even after recreating the SQL databases and executing the DC Directory fix they have postd on CCO I was still having problems and still no Web Attendant. Cisco sent me a procedure on deleting the Subscriber info from the publisher. I forced all the phones to re-register on the publisher while I rebuilt CM from scratch. When I was done the databases were replicated and all my problems were fixed, including web attendant.
Are you getting this error “Installer User Interface Mode Not Supported. The installer cannot run in this UI mode. To specify the interface mode, use the -i command-line option, followed by the UI mode identifier. The value UI mode identifiers...
The below trick might come handy when you have to add a new node to a cluster but you don't have or is unsure of the security password for the publisher. This procedure has been around for ages.
1) Login into the CLI of the Publisher.