Thanks for clarifying that for me, Marcus! I have a quick follow-up question for you.
As you mentioned and I've seen in practice, you can attach a Data Disk Offering to your VM Instance. I haven't looked in CS, but perhaps you know off the top of your head if you are able to have multiple Data Disk Offerings for a given VM Instance. Here is an example: An app like MS Exchange might store its data in one volume and its logs in another. Can I create two data disks for it to use (in the Create Instance wizard, it looks like I can only choose one)? Perhaps the way CS works in this case is you get one disk and you'd create two partitions on that disk within the VM (as in the user would have to do this partitioning manually)? Thanks! On Tue, Jan 29, 2013 at 5:57 PM, Marcus Sorensen <[email protected]>wrote: > You can apply storage tag to both compute offering and disk offering. > > Root volumes are created from a template, and are only created when a > VM is created. They are put on the storage based on the tag the > compute offering has. In other words, when you create a new VM, it > looks at the storage tag of the compute offering and copies your VM's > template there, creating a 'root disk'. > > Extra volumes can be attached to your VM, and they are created via the > disk offering. This model is efficient in cloud because it allows > templates to be small, deployed and backed up quickly, and then extra > disks are used for large file storage. Those extra disks can also be > detached and moved around between VMs. Of course, if someone is used > to the traditional way there's nothing to stop them from creating a > 100G template and just having a root disk. > > At any rate, create a compute offering with your desired storage tag, > and create a VM referencing that compute offering. It should be > deployed on the storage you wanted. > > On Tue, Jan 29, 2013 at 4:53 PM, Mike Tutkowski > <[email protected]> wrote: > > Hi everyone, > > > > I'm continuing to learn some of the basics of CloudStack. :) > > > > I was able to create an iSCSI target from an Ubuntu VM of mine and > enable a > > XenServer VM of mine to see it. > > > > I went into CloudStack and created a new Primary Storage type based off > of > > that storage (by specifying PreSetup). > > > > I then went into Disk Offerings and created a new one that leveraged my > new > > Primary Storage type (now, correct me if I'm wrong, but the way I did > this > > was to use the same Storage Tag I created with my Primary Storage type as > > the Storage Tag of my new Disk Offering). > > > > I later created a new VM Instance and selected a Data Disk Offering equal > > to the new Disk Offering I had created. > > > > This all seemed to work well. :) > > > > Now, I was curious, it looks like my VM Instance (which is a > tinyOffering) > > is running on local storage of my XenServer. How was this determined? I > > looked at the Compute Offering. If the Compute Offering would have had a > > Storage Tag of my new Disk Offering, would my VM Instance have been > placed > > on that storage? > > > > Thanks for clarifying for me! > > > > -- > > *Mike Tutkowski* > > *Senior CloudStack Developer, SolidFire Inc.* > > e: [email protected] > > o: 303.746.7302 > > Advancing the way the world uses the > > cloud<http://solidfire.com/solution/overview/?video=play> > > *™* > -- *Mike Tutkowski* *Senior CloudStack Developer, SolidFire Inc.* e: [email protected] o: 303.746.7302 Advancing the way the world uses the cloud<http://solidfire.com/solution/overview/?video=play> *™*
