Dell EMC PowerMax Management with Demo with Vince Westin

okay so let's talk about PowerMax management again this may be a point we make a break so I'm Vince Weston Technol evangelist for Dell EMC storage field day 16 here to talk about management of PowerMax erase oh thank you all for your time let's get started in terms of general management architecture we have a GUI we have a CLI we have the underlying api's that allow all this stuff to work and then we have a big REST API collection that allows us to connect to just about anything so you've got provisioning replication migration system information performance all available through whatever kind of tools you want and that all ties into the overall array we'll talk through some of those details over the next couple of slides for example with VMware we have very deep integration as you might expect again gee we've got a few friends at VMware so we you have API integration for VII and vasa we've got vivo support we've got vsi storage integration we've got Unisphere ability to see what's going on so you can look at the VMS that are on an ESX server and understand the LUN relationships and understand how all it's not worse we've got and with ESI you get the same thing from V Center right we've got the storage analytics ties in to be realized ops we've got the power max orchestration with V realized Orchestrator we've got our Tia our SRA 4 SRM for us RDF is again we've been doing this with VMware for quite a while login site whatever you need we tie into a whole bunch of things here with VMware so good friends at VMware lots of good stuff to do glad to help you support your VMware environments REST API we've got all kinds of things so this is a collection of resources I expect you all may want to click on slides and go find some links you can start really simple with something like the github repository which is a Python going into Unisphere for v-max to get sample code and all that stuff so he's defined there's also a whole bunch of things in the community MC comm where you can go things about all the rest API pieces and that kind of stuff so lots of things about that if you need any more help feel free to send me an email I can get you in touch with all the management team we help you find all the depth we've got you know sample things sample code we've got folks who can walk you through you know using some of our little snippet devices that can work with us glad to help you getting up and running doing any of that that you want drilling into like the github this is alright we've got things that multiple pieces multiple people that bid about how to use you know ansible or whatever other particular thing they were interested in how to tie this in with Alexa okay whatever you want to do you can tie it in with the rest api and go chocolate alright so it all depends on how creative you want to get and what you want to do with it how you want to play with github we're glad to support that we can also do some interesting things with the again the Python code so you take all this and you link it through the REST API scripts and it comes down to specific calls and if you really want to do let's get geeky about Python we can talk at the level of you know how do you name your functions and set up the calls and do some of those things I'm assuming at this point in the afternoon none if you really want to talk through each of the individual pieces but again we've got samples like this some very definite knows we've got lots of samples like this we're glad to help you work through any of that if you or any of the folks you're dealing with you know want help to go play with this we're glad to help you do it we just wanted to put this up as an example right to give you a sense of how it's going to work cuz we didn't figure you'd want to go through it line by line all right cloud IQ support again Susan's in the back she's going to talk to you about cloud IQ in a few minutes so we have the ability doing the early access program to start tying this in for our raise now and again you'll see that full support coming in like September so you have the code now on the power max arrays when we can do it on v-max all flash we're planning on taking that all the way back to the early beam axis so we'll have you know 76 code 77 codes 78 code and potentially mega 76 guys early we're going so I'll give you all that so great options to see this and then we get a chance to roll up data and so you'll get a chance to see this for your power maxes but you can always see it already see it for unity you've got extreme IO coming and other things and you know I'll stop talking about that let's fire IQ because it's pretty awesome all right so what have we done with Unisphere for the GUI management powers use fear for power max version 9 is the latest and greatest version it has been completely rewritten so that it is now all html5 based so we have the ability to do things much faster much more flexibly we support all the new role based access control options we've got the service level information we've got the non-disruptive migration support changes all in here all the kinds of things you'd want to do for managing all this and again support for you max 1 2 3 all flash all of this being brought forward with the management tools all with this simple html5 interface so before I pick up the interface any questions so I'm actually gonna slide up for this thing and type on a keyboard from that this is a demo edition so it is connected to exactly nothing in terms of real worlds hard work but it has some sample data built into it which allows me to do consistent demos alright so there gonna be some things we'll get into and they'll say you might say how do you do that it's like well if I had a real system alright if I were a real boy I'd be able to do that but yeah today I'm not a real boy so at the home level you can see your multiple arrays hopefully you know you're a good customer you've got 20 or 30 arrays sitting in your data center and you're buying more and that's a great thing we want to make it easier for you to find them all and be able to quickly understand you know how many storage groups you have how healthy are they is your array and compliance what's your you know health score like what your throughput like you know it's a high level view which array should I pay attention to what's going on I can take a look at for example this particular array and it says gee I'm in this array I've got a certain amount of capacity so in this case I have subscribed 114 terabytes of data and I've written to 44% of that right so I've written 50 veidt's or so on the front end in this case of that because I got several snaps and things going on I've consumed almost the same in usable capacity so I'm getting kind of high on that my over subscriptions about a hundred and eighty six percent my overall efficiency is about three point two to one and I can look at things like capacity trends where I can see you know my usable sitting flat on my adding capacity but I keep adding London I'm subscribing more and more at some point I will run out of space and I can look at some historical performance and I can change again use the easy buttons here for for our performance one day to week you know put overviews of what's going on in the box if I'm worried about health I come over here and I click on system album and it would show me the scores and the individual pieces and then if I wanted to drill into like storage groups if I had a service level compliance issue and I might just say hey show me my individual storage groups what's going on with that right and I can jump right into the storage group level and go look at individual storage groups I might say gee you know what's going on with this particular storage group and it'll give me lots of details about it and then I can go in and change things set host i/o limits change the protection I do whatever else I need to do to bring things into compliance or to change other things in fact while we're here let's go ahead and do a demonstration on protection so I can take this guy and I can say gee I want to turn on time finder for this so yep let's go turn on some local snaps and I want to build a new group let's say I'm going to call this biz events one names of the snaps we're gonna do a time to live expiration we're gonna make these guys die after 12 hours all right so it's gonna create snaps they're gonna live for 12 hours right now by default it'll make one of them right now or I can set a schedule and I can say yeah no I really want to recur on with these to recur I want to start them today I want them to start at I can't start them in the past let's start them in the future start them at 5 o'clock and I'll let it go five o'clock top of the hour every one hour so what this will do is this will create snaps every hour on the hour and I'll keep them for 12 hours and so I've got this nice set of backgrounds okay that's what I want for my schedule I can look at this and see the summary of what I'm doing right expires after 12 hours not being done on both sides of RDF I want him to go every hour starting this afternoon at 5:00 yep go right add that to the job list it'll get processed and then I can go back to doing whatever I want to and I can look in the job list and I can see when it's done and lo and behold now I've got snaps going every hour for that that's how much work it takes to build snaps right not really difficult and if you wanted to you could set the the this secure snap flag on there as well but that's in the Advanced Settings and done so what's the warning on the first storage group they're from the compliance perspective he's saying he's marginal he's got a response time challenge again this is a sample data set so when someone was playing with sample data set they played with it the wrong way and it turned it marginal compliances compliance against the service service level service low light so your goal on diamond is around point six milliseconds if you drift over a millisecond half right you're gonna go well over a millisecond or so you'll go yellow over two milliseconds you'll go red right and if you're red then you know somebody come fix me as it starts drifting up of course the array would say well gee I've got these gold things in the array that aren't as important I'm gonna slow them down so I get more performance on Diamond now some of the golds may drift out of compliance at that point but right as long as my diamond stuff is May and I don't care about gold all right so it'll do all that kind of stuff automatically so I can use this to set up typical RDF where I can do already FA or synchronous in this case I'm going to want to demonstrate too many things because we're running out of time here so I'm going to demonstrate s already off Metro and just show you how easy you know we walk through on the slide how easy it was to do nd M and I'll do a quick set of it in here but let's show you how easy it is just to build up a metro config so I say I want to do Metro on this and it says well gee I see already that the only array you're attached to that could read berated ooh Metro because of code levels and all is this guy 647 right there's nobody nobody else in the list - but he's the the target I'm going to pick I'm going to let it automatically pick the RTF group I want to join I'm gonna automatically establish pairs for the new Lunz I'm gonna that I want to use a witness it'll pick the witness automatically because RDF stuffs already established I'm gonna use the same storage group name on the target that I have on the source and I'm gonna use the same time in service level on the target that I have on the source all right so I'm ready to go do that and here's my chance to review it when I add this to the list it will reach into the target array it will create the new storage group it will create Lunz that match my source ones it will do the RDF relationship to them it will copy all the data across and it will copy the personalities across bring them up and you'll be able to match those to a host but they will be identical from a worldwide named serial number all that point of view so the host will just look at them as alternate paths the same one right so it's that easy to go ahead and create a metro configuration the real difference here in terms of this and migrations is this won't do the masking to the host because it doesn't mean you do Metro you may want to mask it to a new host it's not necessarily back to the same one it's not a nice early migration this also doesn't tear anything down on the old one right you're building your relationship not destroying anything so this is just gonna create a new relationship I can't book this in the demo version so I'm just gonna cancel out of it but it shows you how easy it is to build something like a metro relationship I also can take if I want to take a storage group here and say I wanted to do in DM since we're talking about MDM I can say I want to migrate this guy I don't know I'm asking of you are I thought he was one of the head I'm asking of you you can tell I spent all my days when I'm asking for his right so that I know exactly what I supposed to migrate migrate me all right I thought I had three in my head that we're working right and now I'm not getting it so yeah nothing like a live demo right all right so I've got a storage group I'm migrating it I picked my target array in this case there's only one array that's running the 78 code that's going to let me migrate to it so I can pick the DES RP I want to go to in this case the frame has two resource pools so I can pick SRP one both at the port group go automatic I say next I'm going to create the data migration I'm going to allow compression on the on the group and so now here's my view it's going to go ahead and use this mask create this masking view right with this port group to the host group right it's going to allow compression on the start target and away we go right that's what it takes to do the create and then as we said when this gets done all I do is a commit and I'm done so it doesn't take a whole lot to move data between frames right pick the group pick the target frame a couple of parameters got we're trying to make this really easy for customers to be able to pick up data and move it to new frames the real goal is and when you when you buy a new array we want to make it easy so you can take data from many arrays and put it on the new array right as you buy a new frame odds are what you really want on your latest and greatest and fastest array is your most important application which probably isn't on your oldest array that you're going to get rid of so want to make it easy to take the best application and move it onto the new frame and then migrate the others from the older frame to the less old frames right and then pull the oldest one out of the environment all non-disruptive lis want to make it easy to do those asset walls alright so that's it on that let's go back home and look at a few other things we can go looking at let's say yeah let's look a PM wear right now so we have an ESX view we have some virtual centers that have been registered we can go click on one of the servers and go look at the properties write memory and cores and all those kinds of things we can drill into the masking views that are in use to allow that thing to see various ones and what arrays they're from and you can drill into the details on those we can drill into virtual machines and see lots of information about the VMS and the memory and the virtual CPUs and all that kind of stuff we can go looking at performance and see how things work you can look at for example this particular storage group and here's the front end directors and here's the front end ports and here's the storage groups that's a part of and you can really quickly get from whatever your ESX server is all the way down to storage components behind it and understand your performance and know how everything is working so again we want to do this from the storage side the same kind of thing happens on the VMware side with the VMware storage integration tools we have all the hooks into the center and so vCenter can see the same kind of stuff from the array to the array side only from VMware so now when the ESX our VMware admin calls up the storage guy and says hey I'm having problems with ax he can pull it up and say hey okay let's look at the same ax and we can agree to where the problems are and what we're gonna work on it just makes it much easier to use the same terminology get on the same page see all the same pieces all right anything else in particular we want to look at I guess again look at some of the protection pieces so I mean all this information is at the storage group level if you wanted to see like a specific LUN kinds of things can you drill down to that lecture let's see I'll go take a look in the compliance over up into a storage group so I'll take this guy and he opens up and I can drill into individual volumes yeah yeah okay slowly loading the views but yeah but you can drill down further and get to whatever you need to see to from the volume level right how much is allocated and all kinds of pieces about those guys and then you can do things like you're like you can also say I want to expand this guy all right I can just simply take hundred now it's 200 gigs right add that to the job list done right so LUN level my great management's easy to do add lence the storage group just at the storage roof level create more capacity change service levels whatever you need

Channel: Tech Field Day

Views: 6,706

Rating: 4.8666668 out of 5

Keywords: Tech Field Day, TFD, Storage Field Day, Storage Field Day 16, SFD, SFD16, Dell EMC, Vince Westin

Id: lNIOyEZH01c

Length: 17min 12sec (1032 seconds)

Published: Thu Jun 28 2018