Capacity Management is one of the harder things to manage in an IT environment both from a technical side and a policy side. What we can help with here is the technical side of capacity management. Give you the ability to show management in black and white how much capacity is remaining and then it’s up to them to act on that information.
In this chapter we will look at how to use policies and correctly so we can properly do capacity management across your entire virtualized environment.
Logically breaking up your environment
Before we can create some capacity profiles we need to break up the environment. The typical break up would be your usual test/dev/prod. But it can be 3anything; another example would be different clusters with different hardware types that may need different policies applied.
The following steps will guide you through the groups creation:
1) First things first, we need to log into the vSphere UI https:// the below image is now what you should see.
2) Now we want to create some groups, there is a few ways to go about this but what we will do here is left click on the little icon on the bottom left which looks like a bunch of shapes with a green cross. This is shown in the image below.
3) The New Group dialog box should now be visible, this is where you enter the Groups details.
Name – This can be anything you like, but best to make it something meaningful so you know what it is just by looking at it.
Description – Again this can be anything you want, but best to make it something meaningful, If you see the image below you can see its production systems running on HP DL380 servers
Type – This is a drop down menu which should contain Department, Environment, Function, Location, Security Zone and Service Level Objective Pick something which bests fits with what this group you’re creating is. I have chosen Environment which is my “production” environment
Policy – If an existing policy has been created previously that fits the group you are creating please select that, but because we are creating this group specifically to add a policy to just leave it as the default.
Membership Type – What we are creating here today is a dynamic membership type, basically this means we will be using rules that will dynamically add objects based on our group membership rules. The other option is static and you can hand pick the members of the group.
Please make sure that update group membership automatically is checked, this means when a new host of virtual machine is created it will automatically get added to this group if it matches he membership requirements.
Once completed left click next to move onto defining membership
4) We are now on the Define Membership page as pictured in the below image, in here we define the membership for the production group we are creating, this is done with a GUI configurable function, which allows you to configure the rule list for group access. You are also able to manual include or exclude objects manually from the group.
In the below image we have 2 different rule sets, one for all the descendants of an object which in this case is a cluster “virtualiseme” and the second rule set is for the cluster itself. You may notice the “OR” when you select “add a new rule set”, This means that any object that meets at least one rule set will be included in the group. If we were to select “add a new rule criteria” it is like an AND in programing, meaning a group member must meet both criteria’s to be added to the group.
The drop down menus you use to create the functions will differ depending on what options you select. Because we picked string based variable like “name” and “Descendant of” means we have a choice of equals, contains or not equals. If we had of chosen a numeric variable like number of datastores we would then have equals, less than or greater than
Now once we are happy with the group search criteria we can preview the result to see if it’s correct or we need to fix it up. To do this select the preview button on your current screen and you will get something similar to what we have in the below image.
As you can see we have virtual machine, hosts, datastores and the cluster itself in the list so we know our functions must be right. Click on close and since we are happy with the rules we have put in place now click on the next button
We are now sitting on the review page, this page gives us a change to look over everything we have done, You will notice we didn’t touch the always include or never include options. We didn’t need to use them in this case. And I try to avoid them as I like my groups to be dynamic as possible. Explicitly excluding or including is like hard coding ip addresses in code. Because we are happy with the group settings click the finish button.
Tadah you first group is now created, You will see it under the groups menu in the main dashboard like in the below image, You will also notice with that new group selected it now has its own badges to the right which will be grey and have a “?” in it. This is because it will now have to work all those badges out for the new group.
We now have a production group created, you can go through and create all the groups you need to, to represent test and dev or other environments. For capacity management its best to keep it at the cluster level, you can create groups from folders or resource pools or vCenters you’re not confined just to clusters.
When dealing with cluster that have multiple levels within the same cluster you can do resource pools but since capacity is ultimately done at the cluster/host level its best to do a group for the cluster as well and come up with a good middle ground so neither environment level is impacted.
Next up is:
part 2 – Profiles
part 3 – what if?