So I (Drew Tonnesen, @drewtonnesen) am guest blogging again here at Cody’s site. I’ll add my name now in the beginning to avoid the confusion (yes there has been) that this post isn’t from Cody but given the blog URL I suppose it is inevitable no matter what I do. And since I am not starting my own blog which would more likely contain hints on how to avoid going insane with kids (yes alcohol helps) I trust the readers can figure it out. On to the post…
I’ve been working on this new release for EMC Storage Analytics (ESA) 2.0 for what seems like ages now. The first two releases of ESA (1.0, 1.5) supported the VNX platforms. The 2.0 release is the first with VMAX support! As the Symmetrix engineering contact for the VMAX component of ESA, I have had first-hand input and testing of the product, along with putting together a good part of the documentation. This post jumps the gun a little as the product will be officially revealed next week at VMworld but what’s a little pre-announcement among friends? I’ll give a brief run-down of the release below.
As a brief intro, EMC Storage Analytics (ESA) is a very powerful management tool, designed for VMware and Storage administrators, that can see across your EMC storage platforms to provide overall health and risk analytics. It works in the context of VMware vCenter Operations Manager. Basically what it does is provide storage metrics directly from the array via an SMI-S Provider. Of course it has all sorts of great features like integrated and customized dashboards with complete visibility into metrics and analytics. All this is done for the purpose of assisting with problem resolution.
If we think about how you go about resolving a problem in the environment, basically you are doing three things – define normal, identify the problem and identify root cause.
- Define normal – What did my system look like when no one was complaining? What did the metrics look like? Which of the underlying components is doing something out of whack that is causing this problem? You might look at a previous day, week, month etc.
- Identify the problem – As an admin much of the time can be spent proving that the storage array is not at fault even if it’s performing as designed. High response time – yes it’s high because the application is sending more work down or the workload profile has changed. Storage is the last in the stack so blame gets unduly meted out there.
- Identify root cause – This is the last step – get to the root cause of the problem so that you can fix it, but also prevent it from happening again. ESA is going to make that a lot easier.
Let’s take a quick look at the dashboards that are delivered with ESA. There are two universal dashboards and one specific to VMAX. Our first dashboard is EMC Storage Topology. Here you are going to see all the VMAX arrays you have configured (and other arrays if you have those) – we call them adapter instances. You can navigate down each array to see the underlying components – from storage groups to FE ports. ESA is also going to pull in all those related VMware objects. So for instance in the screen shot we can see that LUN 184 has datastore VMAX_ESA_Datastore built on it. Note too that the LUN is in multiple thin pools – the storage group happens to be in a FAST VP Policy.
The second dashboard is EMC Storage Metrics. This is where all the analytics work can be done. Each EMC resource will have a set of metrics associated with it – in this example we are looking at the metrics for a storage group. Each metric can be displayed in a graph with many different options available – from combining graphs to doing historical reporting.
VMAX also has its own specific dashboard (as does VNX) which is done in the form of a heatmap – EMC Performance Overview. This dashboard will cover the main resource types – thin pools, storage groups, LUNs, FE and SRDF directors, and SRDF groups – and provide a few metrics from each one. The heatmap colors work on two different levels – there is a green to red legend for some which represents either usage (e.g. thin pool allocation) or performance (e.g. latency) and there is a blue legend for relative usage across that metric within an array (e.g. total writes). From this dashboard a quick graph of a particular object (e.g. SRDF director) can be generated or using the icons you can navigate to the metrics page to get a full historical perspective of the metric values for that object.
Note that the dashboards are fully customizable – if you want to clone the EMC VMAX Overview dashboard and adjust say the response times for latency feel free. There are also many other widgets besides the heatmap that can be added to yours. We also have a dashboard exchange on the EMC Community. This will allow customers to upload and share dashboards they create:
So that’s ESA 2.0! Look for it soon!
Call to Action at VMworld 2013:
If you are attending VMworld 2013 there are a variety of ways you can learn more about the ESA VMAX release and here they are:
- Hands-on-Lab: Introduction to EMC Storage Analytics for the VMAX
Interact with the ESA for VMAX in the EMC Hands-On-Labs! For information on the labs check out this link: https://community.emc.com/docs/DOC-25882
- Session: VCM5539 – The Missing Link: Storage Visibility In Virtualized Environments
Definitely attend Matt Cowger’s (@mcowger) session on ESA (among a few other things) to learn the details of this new product. Find the abstract and schedule here: https://vmworld2013.activeevents.com/connect/sessionDetail.ww?SESSION_ID=5539&tclass=popup
- EMC Booth Demo: Check out my video demo at the EMC booth of ESA for VMAX
See this demo (and many more) and discuss with the experts in the booth what it is and how it can help you.
- EMC/VMworld vPass Launch Page
Of course for all things VMworld and EMC check out the EMC vPass. Worth it for all of those attending and not attending VMworld this year. Find it here: https://community.emc.com/community/events/vmworld