Upgrading to Scense 8 might seem scary after you’ve seen the upgrade notice. It tells you that Scense 8 is not compatible with any of the previous versions and your workstations will be forced into local cache. The fact that the Scense setup offers an in-place upgrade promises at least some compatibility…
After having completed the upgrade process myself I can assure you that your Scense system and its contents will be just fine and everything seems to work just like before the upgrade, so the upgrade notice seems a bit exaggerated, or does it? If you go at it without any planning and just perform the upgrade; then everything will happen just like the upgrade notice predicts. So, a little planning seems to be in order.
Before rushing into the upgrade it’s probably good to think about the best approach. The notice tells you that your workstations will go into local cache and you should update these workstations immediately after the upgrade. While this may be feasible in some (smaller) environments, it may be too radical in the larger ones. Although the local cache will allow users to continue using their workstations, probably without even noticing that they’re running from local cache, the users will very likely be unable to install or repair applications and the administrator will not be ‘in control’ during this period. Upgrading all workstations with the new Scense 8 client will probably take considerably more time in a large site than doing the same in a smaller environment. So in large companies the local cache period will be longer or perhaps too long.
While it’s possible to go big bang with the upgrade, it’s also possible to take a more subtle approach.
The subtle way to upgrade to Scense 8 is by creating a shadow Scense environment. This is particularly interesting when your Scense server is a virtual one. A shadow upgrade is done relatively easy by copying the Scense database, cloning the Scense server and making absolutely sure the two are connected properly and run independently from the production Scense system. Once the shadow system is up and running you might want to freeze all modifications on both systems, so you don’t run into surprises after the upgrade.
So now we have an identical copy of our production system which we can use to do the upgrade on without affecting the production system.
Just do it
At some point, after all preparations have been done and (very important) all prerequisites have been met; we just have to start the upgrade. Early in the upgrade sequence, the Scense setup will ask you to ‘disable and shutdown’ all your Scense Engines. In the big bang scenario you would have to do this for all your Scense engines including the ones in a (network load balancing) cluster. In the shadow upgrade scenario you should, of course, only disable and shutdown the Scense engine(s) in your shadow environment and leave the production ones running.
The Scense setup will now modify and convert the database, configure the new web services in IIS, etc. After the upgrade has completed, it’s time to configure a first administrator workstation with the new Scense client and –explorer. You can do this by installing the Scense Update Manager on the upgraded Scense server and push the new Scense 8 client upgrade job to your workstation. In the big bang scenario you might want to set the new Scense 8 update job to be the new default mandatory job, but in the shadow upgrade scenario you should be very careful not to push this job to all workstations…
Once we have our first admin workstation up and running we can start testing and perhaps upgrade some more workstations to the new Scense 8 system. At some point, when we established that everything seems to work just fine, we can start planning and executing the upgrade of larger amounts of workstations. Make sure your Scense back-end is capable of handling the larger amounts of workstations. If you were running Scense in an (NLB) cluster because of your large amount of workstations, then you might want to create an IIS cluster for Scense 8 before you move all of your workstations to ensure performance.
After upgrading all or most of the workstations, we can start exploring the new Scense 8 features. Runtime patterns can be made significantly simpler and faster by making clever use of the new scopes and criteria. Now we’re also able to configure and use the location system. We can now easily route the Scense engine communications through routers or over the internet etc.