March 24, 2015
Atlantis USX 2.2.1 Now Supports Volume Name Templates
Atlantis USX 2.2.1 is a maintenance update to 2.2 that has a number of fixes for customer-reported issues, and we have also introduced a new element in the naming of the USX volumes. This enhancement could be confusing to people that have worked with USX before, but will feel very intuitive to new admins. This blog post is to explain why we have introduced the change, and how it works.
In previous versions of Atlantis USX, the name of the Volume VM and the name of the datastore for that volume were defined by the admin at the time the volume was created. This is logical and easy to understand for simple deployments during a PoC. However, in larger deployments with multiple volumes, you would naturally introduce HA into your Atlantis USX deployment. After HA events have occurred it is possible to be in a situation where the names of the datastores do not match the name of the Volume Virtual Machines.
This is because the name of the datastore really is tied to the volume “service”, and not to the name of the virtual machine where that service is running at any point in time.
The following visual representation of a relatively simple USX deployment with 2 Volumes and a shared HA volume makes this hopefully a little bit easier to understand:
And the same deployment after Vol01 and Vol02 each have an HA event in succession:
Now it’s clear that the volume VM names are not tied to the datastore names anymore. Picture this with more USX and HA volumes - your feedback has been we should make this easier to manage.
Introducing Volume Name Templates
USX 2.2.1 is addressing this issue by explicitly splitting out the management of the Volume VM names and the Volume Service Names. It introduces the notion of Volume VM Templates that determine the Volume VM Name when you create a volume, and allows you to specify the Volume Service Name when you create a volume.
Screenshot of the new Volume VM Template creation dialog:
And the Create Volume Wizard:
Once the volume is created and enabled with HA, you can go to the manage volume page to see in which Volume VM this particular Volume Service is active and which Volume VMs are available as HA failover VMs:
The ‘Active’ indication will change when there are HA failover events to reflect in which Volume VM this Volume Service now lives.
Concluding, the visual representation of the new management model is now:
Hopefully this is clear now :)
Any questions? Then check out the forums.
Current rating: 5 (1 ratings)