Lync 2010 Installation with backend SQL Cluster failed login

I started installing Lync into our production environment after having installed it on a test lab, the only thing that differs was in production we use a SQL cluster as backend. In the test lab I have used a single SQL cluster server as backend.

I created a topology and tried to publish it, it immediately threw me an error with failed login for user DOMAINAdministrator. After searching for some answers via Google which as always led my in the wrong direction thinking it actually was an error on my SQL cluster.

The only thing you need to double check is that on all nodes the virtual sql instance you have opened the SQL Browsing Service (UDP port 1434 by default) and allow for remote administration in the local firewall and all other firewall that might be in between.

Continuing the search on Technet reading about Configuring SQL Clusters in the Lync Deployment pages I found the correct powershell commands to use:

Install-CSDatabase -CentralManagementDatabase -SqlServerFqdn sqlfqdn.mydomain.tld -SqlInstanceName mysqlinstancename

Now i got somewhere, still getting an error but this time it made more sense: error_no_db_drive_access

This seemed to me that somehow Lync was trying to setup the database files and logfiles for the SQL instance on non clustered disks and i decided to alter my powershell command with:

Install-CSDatabase -CentralManagementDatabase -SqlServerFqdn sqlfqdn.mydomain.tld -SqlInstanceName mysqlinstancename -DatabasePaths X:lyncdatabaselogs,Y:lyncdatabase

And Halleluja! it created the CentralManagementDatabase! After that I continued to publish the topology. And for the other SQL Databases that needed to be created I choose the advanced button and instead of "Automatically determine database location" I selected: "Use SQL Server instance defaults".

Comments (2) Sep 27th, 2011 @ 9:03 am CEST

Synology ds710+ pxe server with ping and dd-wrt

This weekend I decided to put in place image capture and restore functionality in my network using all the pieces of hardware i already owned. Reason is that I do not like it when re-installing Windows I have to do updates etc nowadays you'll lose an entire day re-installing. For imaging I use ping. Alternatively I will see if I can add unattended in the pxe menu and rollout unattended install's of my OS's. Will blog about that later.

I will write down the steps I have taken to achieve my goal:

Continue reading Comments (1) Jun 14th, 2011 @ 8:56 am CEST

Remove Missing or Failed status VM's in SCVMM

Sometimes when SCVMM F***s up and you are left with double entries of VM's seperated over your cluster one is OK and the other Failed or Missing. Here is what to do:

Continue reading Comments (1) Aug 27th, 2010 @ 8:40 am CEST

<< First < Previous [1 / 3] Next > Last >>