Recently I posted a link to my new video on Tintri VM Scale-out. I thought it would be a good idea to explain the different functions and capabilities in the blog so I can go into a bit more depth, cause that's the sorta guy I am!
So, where was I? Oh yeah, Tintri VM Scale-out!
Here's the front page of Tintri Global Center or TGC that we all know and love. If you haven't had a look at it, I highly recommend checking it out!
Sorry for the little images, if I made them bigger they would go off the page, so if you want to see them in more detail, just click on them.
I won't go into detail on this image cause I'm going to focus more on VM Scale-out. Maybe if you ask reaaaaaaaaaaaaaaaally nicely, I'll do a TGC in depth blog!!
To get to the new VM Scale-out tool, just click on the Pools tab.
Now remember, VM Scale-out allows you to seamlessly expand your Tintri VMstore infrastructure using analytics, predictive modeling and proactive resolution!
Woo Hoo! Check out all those cool numbered bubbles! Yep, I'm going to discuss each one of them!
Unfortunately I have to explain somethings out of order cause I'm a ding dong and numbered the bubbles without really thinking about how I was going to write the blog! Ah well.
#12 +: Yeah I know, 1 comes before 12. Ugh.... To add a Pool, just click on the + sign!
All you've gotta do is name the pool and select the VMstores you want in it. If only everything in life was this easy!!
#11 Pencil: Hey, let's continue on the backwards trend! The pencil edits the pool. Let's say you want to change the name or add or remove VMstores from the pool. No problem, just click on the box to add or X to remove. I know what you're thinking, "It can't be that easy!" Well my friend, I'm afraid to tell you, it is that easy.
#3 VMstores: If you click on this, it will take you to the page in TGC that shows details on your VMstores.
#4 Virtual Machines: Click on this and you'll go to the TGC page that shows in depth metrics on your virtual machines in that pool.
#1 Recommendations: This is the one I really like! VM Scale-out analyzes your VMstores, virtual machines, IOPS, Throughput, Latency, Performance Reserves and Space Remaining. From all this information it comes up with recommendations according to how the environment is doing. Let's expand the recommendations and take a closer look. How do you expand it? Just click on it. Want to execute the recommendations? Click on Execute.
Things are broken into three categories, Current Issues, Recommended Actions and Outcomes. A LOT of stuff is clickable, so let's click around!
In the current issues we can see our Tintri 820 is running out of space and is experiencing heavy load. Want to go to that VMstore directly, just click on the name of the VMstore. Want to see more information about the current issue? Just click on View Details.
We can see we have a 3.1% chance of running out of space on the VMstore and that it's 62% full and how much it's increasing per week. Pretty cool huh?
Let's move to Recommended Actions. VM Scale-out is recommending to migrate 41 VMs to solve our space and heavy load issues. If you want to see what VMs those are or if you want to tell VM Scale-out certain machines shouldn't be migrated, you can do it!
I've clicked on the 41 VMs and here's what I see. I can filter, search, sort, all kinds of good stuff, but what's really cool is I can tell VM Scale-out whether one of the selected machines should be migrated, don't migrate right now, not to the selected VMstore or never!
#2 Notifications: This one is pretty cool. While VM Scale-out is checking space and performance it's also checking for issues in the environment and not just the VMstores! Remember, Tintri is VM-aware so it's checking for issues there too. For VM Scale-out to successfully migrate machines from one VMstore to another the datastore paths need to match. Here I've got a notification telling me that my Tintri 880 has a unique path and problems could arise from that.
#3, 4, 5, 6, 8 and 9 I talked about in #1 Recommendations.
#7 Latency: If I click on Latency, I can see the overall latency of all the VMstores in the pool. Here we only have 0.6ms of latency and it's coming from host and flash. The cool thing is you get an over all understanding of where the latency is coming from in your environment. This really helps us understand where the latency is coming from and hopefully stop it before it becomes a problem.
#10 Migration Rules: Remember in recommendations where I could delay the migration of certain virtual machines or choose to never migrate them? If I want to set up rules like that even before VM Scale-out does it's analysis, I can do it here. Here I've got a virtual machine called r_grow_space1. It shows where it currently lives and I can set rules to never migrate it or never migrate it to a certain VMstore. This is very handy if there are virtual machines you just want to live on a certain VMstore.
Well, I think that's all the numbers! Neat stuff huh? I think so!