Tag Archives: wcs

Upgrading to WorkSite Communications Server 8.5SP2

I first saw a demo of WorkSite Communication Server 8.5SP2 when it was introduced a year or so ago. Since then its had some missed deadlines but finally was released in GA a couple of weeks ago. Partly the reason for its delay however was because of the many architectural changes in the software. The underlying code has been completely overhauled and the operation of WCS 8.5SP2 has been changed in a number of ways.

WCS 8.5SP2 has a number of key features all rolled up into one release

  • Greater integration with Exchange 2010 CAS
  • Load Balancing across multiple WorkSite Communication Servers
  • Mailbox Sync/Mailbox Agent
  • Numerous bug fixes

 

If you are moving to Exchange 2010 and currently have WCS SP1 installed you may want to check my earlier post for further information.

In this post I’ll focus on a couple of the features that come with SP2 and share with you some of the useful tips I picked up along the way.

Exchange 2010
One of the eagerly anticipated feature of this release, especially for larger firms, is the deeper integration with Exchange 2010. In SP1 EFS configuration was based around filtering on the WorkSite database/Exchange server pair; which was fine until the number of filing folders in a single database increases to unmanageable levels (how much time have we all spent waiting for that EFS pane to refresh praying it won’t crash??). Even then you could have explicitly list your Exchange 2003 mailbox stores (until of course the number of folders per database per mailstore got too large!) to filter further.

In Exchange 2010 Microsoft recommend you use the CAS alias for all external apps, meaning you were confined to using just one server if you couldn’t split your WorkSite database config down further. Autonomy still allow you to filter Exchange 2010 mailbox/database, however with the introduction of Load Balancing, this is less of a necessity.

Load Balancing using iManage Clustering
The core problem with running multiple WCS SP1 servers was that they all ran independently and it was up to the WorkSite Administrator to set up each of their functions according to infrastructure resources and business requirements. If one of the servers failed there was no mechanism for the processing to fail over to another server. In SP2, Autonomy require an installation of iManage Clustering, the same service that clusters middle-tier servers together. This works pretty well, you can now have multiple servers processing the same setup together by the efficient Clustering service.

I would advise you spend time to get to know how Load Balancing works. The recommended setting is to enable Automatic Load Balancing and let the system run on auto-pilot. This works well, on the whole, the users are split evenly across each server with the ValidUsers.urs file dynamically updated whenever there is a change in one of the nodes across the WCS cluster. Bear in mind once you click on Connect, all the options are greyed out and you will no longer be able to make any config changes. Just as well really, I agree with Autonomy this is necessary to prevent any unintended user action that may disrupt the process.

You may choose to run the manual Load Balancing, which is mandatory when filtering users according to WorkSite group. This is definitely a gem in the release; as a WorkSite Administrator I’ve come across a situation many times where users complain folders have not been processed for filing or some other problem. In SP1, to troubleshoot the problem meant to let MarkingWork complete a run and then trawl through the logs. However the logs can get chatty (in SP2 unfortunately the situation has gotten worse) and it is difficult to filter threads for a particular user. In SP2 all you have to do is add the user(s) into any WorkSite group, add the name of your group under User Group Name and away you go, MarkingWorker now only targets the users you want. Simple, but brilliant.

SP2 Tips

  • Use the new functions available to SP2 to go through your infrastructure again and see if it needs to be tuned. Especially if your firm is over the intial period of introducing 8.5 client, you may want to re-evaluate whether you still need to have an aggressive polling interval.
  • It also might be an opportunity to do some housekeeping & go through the EFS console to remove some of the folders marked Failed as a result of the user filing and then deleting them from Outlook. If in doubt, reset the folder and attempt to process again so you can be sure its ready to be deleted. Alternatively, you may want to check the status of the folder from Outlook.
  • Consider using User Group Name, should you want align your EFS configuration to target particular users more than others.
  • Be aware of how big the logs can get. Currently there is no way to tone down the logs from Verbose. The available setting only allows to tone down to Information or Error in the EFS console view.
  • Although you may not be using Mailbox Sync and have it set to Manual or Disabled; you still need to review the settings you have set in this node in EFS. In particular, notice how any WorkSite/Exchange settings change you make also get replicated into this window (under the Exchange Servers/WorkSite DMS area). If it hasn’t replicated, click away and then back again.
  • Clustering the EFS servers will require a DNS alias first, so get these in place before you begin your upgrade. If you aren’t using clustering, you are still required to input the hostname in the Cluster Name field.
  • When using Load Balancing, note how EFS generates files in the (hidden) ProgramDataAutonomy folder until LoadBalancer has finished at which point the ValidUsers file is placed in the Config folder. This file is the heart of the Load Balancing operation and initiates & controls the process.
  • Using Automatic Load Balancing means each node in the cluster has a number of users held in the ValidUsers file for that node to process. This will not get refreshed until there is a connect/disconnect on that or other nodes within the cluster. Which is fine, until the time you have new users added into your WorkSite databases who will soon start to create Filing Folders. These new users will not get added until there is a refresh and the ValidUsers file is reloaded.

 

What do you think of WCS SP2 so far?

 

Mubashir Mian is the Lead System Specialist at a major international law firm. His LinkedIn profile can be viewed here (http://uk.linkedin.com/in/mubashirmian)

Share