After having some time to spend with Tintri datastores I can honestly say it is the most administratively liberating storage that I have ever managed, and I use the term managed loosely. First, the installation of the storage array was quite easy. Take it out of the box, plug in your power, management cables, data cables and turn it on. That’s it for the physical part. For the logical side, you have to give it an IP address and connect it to vCenter. Yes that’s it. No LUNs to carve, no extents, no special setups. Not many storage providers can come close to that kind of setup ease.
The hardware is also surprisingly simple. The magic in these storage arrays come from an intimate knowledge of VMware and what the consumer needs from it. That is mostly encompassed in the firmware, logic and higher functions of the array that you never have to see. You get an all in one unit that has redundant controllers, management interfaces and a mix of hard disk drives and SSDs.
Once setup is complete you are ready to provision Virtual Machines. Because of my experience with the storage solution, and make no mistake this is a solution not a storage system. I would recommend that you first make a template with your company specifications. After which you may create as many as 1000 (or more) vm’s on my model with a couple clicks via the Tintri interface or using the vCenter web plugin. It’s pretty simple. It creates all the vm’s with almost no additional space, sort of like a VMware linked clone. The only space consumed is delta changes from the base template. You may also use snapshots as a template or any vm on the storage.
If you are a power user like me, then you will be glad to know the solution is also feature rich. For instance, if a user or manager mentions that a system seems slow, using the dashboard you are able to see latency not only for the array but network and hosts as well, you can even see cpu rdy for that vm.
What most administrators will like is that with the logic the solution is built on, most if not all data is served from solid state drives. The hard drives seem to function as storage only. Once data is read or written the first time it is kept in flash. The chance of it moving out of flash I extremely slim. There is no block sizes to set or vmdk’s to assign.
For my use case I am utilizing the solution as an immediate recovery point as well. With the ability to schedule snapshots of individual vms or the entire array, I have opted to do 7 day revolving snaps of the entire storage. What this means is that every day at a time that the administrator can set (I have 12 noon and 6 pm daily) the solution will take a snapshot of all the vms. While you may think that this may interrupt production when used at these times there is no freezing when this happens. It is completely transparent. If a vm was accidently deleted, corrupted or could not be used for whatever reason then a snapshot could be used to replace the vm within seconds, yes seconds! No need to run to a backup product with would increase your Recovery Time Objective. I use this snapshot scheduling along with my normal backup for greater resiliency of the environment.
I have also found that the snapshotting and immediate vm creation come in handy for development use. When a member of my development team needs a copy of a production vm I am able immediately give them a copy that uses almost no additional space. This is seen not as a linked clone in vCenter but a full copy. The only thing that gives away the magic of what happened is the system stats that will state the vm is using less than one (1) MB of space when it is created.
There so much to mention about this solution. I think I will do a technical deep dive at some point for the uber geek in most of us.