Dell EMC Unity Unisphere and CloudIQ Demo with Wei Chen

the wait so why don't we talked a lot about unity being simple you need to being unified so if you want to show us the dashboards and some of the church capabilities out there so let me pull up unity here so this is what our units typically looks like html5 days pretty modern simple you need to be used the design of unity as you saw is about simplicity racially-motivated the GUI was also easy and easy and to use for everybody so as a storage I mean you know I'm pretty busy a lot a lot of time to look at my system all the time so when you first log into the EVGA we give you a dashboard kind of gives you an idea of what's going on on the system right so there's a couple of widgets here dodgy what the hell has to be performance in commands like this any idea what's going on and you can also customize this if you want so if they wanted to add additional widgets we do have more widgets here you can click on add and it adds it right there just like that let's say I see an alert here that you know it's kind of new it's critical and I want to take a look at it I also have the ability to click on this and I'll be me to the lodge page tilted a particular scene see what's going on exactly on the system at certain point in time then listen to this give you quick tour the GUI we do also have a system view page which is the carousel that meant that akashic mentioned earlier before kind of gives you a visual representation of what the system looks like on the GUI so if you look at the back of the system this is what you would see and then if you look at the GUI this is what you would have so it gives you information about the status of the port if the puts up you click on it so give you the MTU size to speed all the information about the health of the systems on here you can spin this around you look at the top look inside the system see what's going on inside this is a faulty component you'll see that here with the span modules theme the CPU the battery all that stuff inside each one we BSP's as well that goes to the front get all my drives so in the front we have 25 drives here again I can click on one of these drives get more information about that particular drive what speed it is what capacity it is stuff like that and again if there's a fault here the drive would be highlighted in orange so that way we can see which Cyrus vaulted matching up the system which would have the fault led on as well but quick tour the of the the GUI so let me go ahead and show you how to provision a LAN as an example so everything in here is widget based it kind of wants to do specs everything you need to do in order to provision something so I'm gonna go ahead to create a new line give it a name I'll call it storage field a there's a couple of options here select the pool size for ulong you do whatever you want you can do it in a thick you can enable data services the compression the host i/o limits or basically QLS instead of any one degree reservation sure host access is next figure host access instructions host you want to give access to so the the definition of the single or double parity is that the storage pool level to the who level yep yep it's when you create the pool yep so when you create a lotta the filesystem doesn't have any options for that email contained in here but you can have multiple pools oh yeah yes I'd get migrated on line between one or level the next step is detection right so I also have an integrated schedule I didn't see any affinity to an SP a at that point no we decided we assign automatically about okay yep so keep it simple right this is the satellite gradual observer generally so we have an integrated stack scheduler its unified you can use the same schedule for file and block you busy check the box and you can fit your own schedule you can customize it and bixie this one says every day at 4:00 a.m. create a snapshot and keep it for two days as an example so they're able after limited number of snapshots I mean 256 provide a spot object for auto resource yep so every line can I haven't 156 next step is application so I want to protect my lawn by replicating it check the box simple as that and at the last page gives you a summary right so everything we figured the sides of the London aim of the lawn give a description if it's thin thick compression QoS scoff shots so many so many lunch can you support I mean 20,000 so depends on the system yeah I got on the high end it's like six thousand six thousand dollar system yeah I have some I think we got a number how many vowels can you support that's in a 10 mm so I think we're on 50,000 at the UH nasty sound desert yes and is that redirect on right - snaps reader that right Greiger's and you didn't show so in the replication you do have an option to set up or they synchronous replication as well as synchronous replication - I think we have already always had it from day one right so I'm going to finish this GUI finishes wizard so if the jobs running I enclose this buddy run in the background a lot of wait for it to finish which is pretty neat voila while that runs I don't go ahead and glue the pools page so you said before we'd have multiple pools here I'll go in one of these pools but can almost it take to create one hundred terabyte one well it's thin so we're not allocating anything what a fine creation with its own by default or like that it is ideal Oh Jess okay so you don't have to create a simple right there's a checkbox for thinner things so you don't so that that's an interesting question actually we still have people they want to reserve this sort is the right amount of storage - not going out of space so we offer the option it is allocating all the space so let me show you I'll go back to the crate 188 there's a didn't box then check box right here alright dinner take you check the box and you get what you want so most of our space reservation like we're directing right so so going back to the pools page right so the loop my pool property take a look at my pool I'm about to stop creating down here so I'll go to usage and then go to storage resources I can get a list of everything that's created off of this pool right so I have VMware VMFS pated stores I have nat servers i have file system to have Lunz consistency groups have V vaulteq data stores - all this from a single pool so it's completely unified from that aspect between one pool we create everything off of that pool well obviously what the consistency group with the group of runs it's a group of lines exactly so we make rusticity we have authoring rights as well so meet a snapshot a constituency group you replicate a specific group in that way you have a consistency across beam type and you delete a long you delete a file system storage goes back right if you do it the delete launch or delete a stop you can click the button here and a lot of spaces we clean back to the pool and because again it's unified pool it can be used for anything else so let's pull very nice so so one of the things we talked about since we're on the topic of snapshots we did talk about about snapshot mobility to remote and snapchat mobile you to do cloud you want to show them a little bit about about how that works yeah so let me go back to the line I just created this is my lawn sfd so when I when I configure this one initially I enable start our schedules as you can see here from a snapshot schedule that's my expense times our schedule I also enable replication as well we replication we have the ability to replicate scheduled snapshots so basically anything that's created by that scheduler that's integrated on the system will be replicated automatically right so that's pretty neat and then I also have the ability to do this manually if I wanted to so I've wanted to take a manual snapshot you know go ahead and create a new snapshot simple at screen plus it gives it a default name okay and the snapshot is here now I can just go ahead and click replicate and they replicate it for me and the cool thing about this is I can also have a different retention policy on my replicated snapshot so on my destination site I can keep it for a longer period of time or I need to have you know different options here for whatever I need I'm going to keep it as a statement it keep it the statement I wanted to as well okay oh I add that into an existing volume obviously that first nap so there's going to be all the existing data right so there's a full replication it's not starting from that point on time so a snapshot is going to be pointed based for the primary source right where every source is already replicated so when you replicate a snapshot would be just but I do know that finally replication exactly okay yep yes and then if I want you for selecting which of the snapshots to replicate in which not I mean it's that flexible right yes yes you'll have to roughly all of them you can select certain one you can do whatever you want so it's flexible like that right the best part of it is that it's quality based so experience just set up the policy once and it's okay everything down here or you can have a 20 min frame polish and apply them you can cool whatever you want right and apply them as the autumn you create and for the ones that are replicated automatically based on the scheduler you also have the ability to customize the retention policy as well as you can see here so you do have a lot of flexibility with that and you are all managing all of this from the one one slide right one one unit sphere right so login to the remote side or anything yes so the UV sphere GUI is also he defied so we saw the hard way beautify the pools unified the GUI is also you decide I can configure ones here I can configure my file systems my VMware both VMFS NFS and also be vaulteq data storage from the same HTML why don't we we showed them how to move to a remote why don't we show them how are we moving snaps to the cloud okay so for that I'm going to show a video that we the vivos block block in files and block and phone yep so we've all can be accessed through ice cozy primer chain of a block and info file we use NFS all right but quickly go to this yeah quickly go through this so this is a video that we made for WMC world so I'll show you this and show you how this works so first we're going to go ahead and you know copies of data onto a lawn in this case it's just a video so we have a lot here it's already created position to my Windows host copy some data on theirs after there so we have something to work with go ahead and look at the video real quick and the next thing we will do is jump into the user cooing right so after this one going into the you said GUI and we're going to create a snapshot just like I did before so go to the properties of the lawn participants data create a new snapshot simple as that and then click OK give it a name then click OK now after that I'm going to go into my cloud management GUI right so this is the clouds here in the pint that we talked about before so CTA will log in here with the configuration real quick in here we have our unity systematic we also have our virtual team cloud added which is it would be our repository to tear to any plans to integrate so I don't have to go to a second UI to do these yeah we have plans as HIV we had a plan for everything has everything wrong I think so we are scratchy to integrate these technology yeah but as I explained before the flexibility that this gear us today because it solving a lot of use case and a lot of people they can move but I just like to see that integrated even if it's a PIL little worse or if that'd be great okay I agree so when you create a policy here let me say everything's policy based so this is all specs ability here we can get the policy based on things like patient time modified time and size name many times that for this particular demo we're going to do creation time greater than one day just as an example right basically any snap which is more than a day or you are going to move it to the clock exactly we are give it with archives out to virtustream so that's just a simple policy but if you do a lot of different rules if you wanted to be creative with that and we do support all that as well what I'm going to do is create the job right so we do have an integrated scheduler in CTA which would run the job automatically so you do it based on you don't certain create a time you could do basic capacity anything like that so that's another thing to the cloud management GUI the CTA good you just only use it one to set up the policy that's it it's not like something you have to go back and pattern every time they pick it up and forget it right so now we want a job right so the jobs running now that we have that the snapshot archive to the cloud we can go ahead and delete our lawn so just to show you how we store works real quick when you delete everything off of our primary unit view system and we'll do a quick restore of the snapshot back to the you are basically dead especially dead right so we do a rescan from our host and our alums gone so we have no data so Michael so disaster strikes now we need to recover based on a snapshot that we archive to the cloud right so again it's just as simple as walking through the same wizard saying which lung do you want to restore to which snapshot do you want to restore that's all catalog here and click the one that you want and then you get to go so once you went through this this is the way you want to restore to this is the unity system in that particular lawn I want to run it now and then finish so once we run it we do a rescan on our host and we'll see that the London magic will be back based on the snapshot restore and we can open it and see our data still there so so obviously is not an operation that we suggest you do everything area yeah but if you do that you have some problem but as you can see I mean you can restore very easily beyond the cloud and they snap in the cloud as I explained they are compressed and and they can be also protected by Ingres because you are asking about the snapshot limit like as I said when you move it to the cloud you are no longer governed by the limit of the system right that 2:56 limit is only for local right so you could have to have physical you could have to basic remote and you could have been finite and you have crazy long with entity image - that's right right so so so never delete a snapshot ever right that's the story all right so that's great yeah we talked about snapshot using for backups now could you also show us a little bit about the snapshots being used as forth in clones yep or for sensitive analytic yep so so going back to my LAN here right so my sft lon click on that click more actions create a clone right so this human created thing clone I can create a based off of an existing snapshot or kinsey say create a new snapshot based off the latest data not to do that for now so this is test dev and again I think was treated as a primary resource a first-class citizen if you will so we do have a lot of options here as far as data services that could be applied to the thin clone right so you do have QoS here you can give it a certain limit let's say give it Mac iOS per per gig 100 and then we'll also give it a bandwidth limit as well I have its own host access completely separate from the actual one so I can give it to a different host for textured purposes it has its own snapshot schedule so it's kind of like the same widget that you saw before again if you know how to create a LAN or a file system it's going to be very similar great replication and then summary and you're done I think I want to show now this yeah so so they did a great job so really I was very happy so I sent him on vacation to Hawaii but he's not sitting in Hawaiian beach right and then we were we ran out of capacity right and the things falling the way up hey what's going on my so so in bomb in Hawaii right I'm on a tablet right I access to my cloud IQ GUI to see what's going on I don't need VPN not only anything like that I get locked in the GUI and see what's going on on my system because it's all with cloud-based all of the data is being pushed up the cloud IQ automatically based on esrs I can look around my my tablet here and see what's going on alright starting to browse around here with my systems and things like that I'll do it over here so you guys can see it three plans for like an iPhone app it actually works on the iPhone by the way by this the only thing is valve area where it is in the browser it was we are not up the app yet but we have an intern of the year we have plans for it so log in tech I like you this is my dashboard again gives you kind of at a glance what's going on with the system you see the different health scores we'll talk about that some more there's some alerts here and then I see one alert hey that's the pools running out of space right so you click that now tell me that to use percentage of this of this pool is architect bull the time the bull which is kind of a trending that we have based on the sort of circle properties of the pool how fast has been growing that will tell you when it's me full the way now is already full I can see information like that right so like what's the system page I can see I have a bunch of different health scores here right so a couple systems are healthy hundred I have a test test test that's just about 70% helps poor and there's an issue here and we had a five different systems right right so these are five different systems this one's in Cambridge this one's in Boston just one in Hocking pain so it's going to spread across or W graphical regions right so I typed up systems happens to be in favorite happens to be a unity VSA up and I see that the health score is actually cooki down in Spartan category so this so nobody not only managing the unity old flag but also some words and the VA say all of them from the singles in spring all the vertical all slivers of unity exactly oh so the health is focused on it's a part of the category so it could be system health it could be configuration issues yeah avid capacity issue here which is negative 30 on my health so difficult it is to take a look I can go down into the system and I can see there's a capacity issue here that says my crucible it was every alert that you get that affects your health score is also an Associated resolution so if I click on this it's going to say well consider adding some more drives to the pool to get out of this situation or alternatively you can also migrate some of your data off of the pool that we have books based so some some suggestions here get out of this issue are delivering snapshots or something like yeah that's also an option as well we give you a suggestion of what how you gain back the points that you long yeah right there and this is the key things that the is it's like your FICO score way is dynamically and also you can be aggregate at both system level right all right and I can just gives more details about my system as well so if I can go into the storage I can look at my lung my data stores and all eyes Megan's here and we can show our capacity as well as performance all of that for every system across here across the world you can see all of that specific tabs here so capacity to have a look at the breakdown of what's being used for what company how much my class fees being used for lunch how much you speak so faster stones down here get a breakdown on my pools I can see this one's full already but other ones can be pulled within a week so that's gonna be something I need to take care of pretty soon and you can you go down further so it's really into processes to find out file types to get more so will collect data about the file type we select the data what apartment some last before it was my security concerns out there will for yes exactly yeah because you don't want to see there's only like data about customers data it is the things like names sizes stuff like that yeah last thing about here in this particular page is performance rate so I can actually look at the performance information of a particular system to see how how many I ops they had SP utilization bandwidth is like that let me go into different system will quick disaster recovery for example it's also anomaly detection on a performance page right so basically what it does is actually not here but basically what it does is it learns what's normal for your environment over a period of time and if anything goes above or below that range that alerting based off of that anomaly as well so you can receive all soldered directly so we establish a baseline everybody Asian from that baseline we're going to learn this is our first into phase zero phase one off of machine learning rates or trying to sort of predict so that's basically predictive analytics assessment this constant we're trying to predict based on your past where well what type of issues you will hit not just from performance perception there's also different views here so I force you to help spur view in all those changes to the forget about the past give you to look at your efficiency right so how much space is use how much is free how much space by saving from using things like thin snapshots compression stuff like that and all the girl raw efficiency number is also posted here as well the other things we track in the including queue is the web level so we show you the we're leveling of the disk yep and we proactive replace the disk if we get so we tell you your class size just don't worry gate baby can you show that yep for Game one which shows the Sony goes by one of these I think we don't know we ever here so it this one's kind of empty right now but basically if you go to the pool and you go to Lee drives view there's a little bar here I kind of show you the fuel gauge how much you have left as far as endurance is concerned we have a stress this one's right out that'd be so because it data doesn't have another slash I'm curious if any you said there's some proactive analytics involved do those automatically feeder guard assume monitoring function at EMC yes any MC says very we're seeing significant degradation and so let's do an example with the we're leveling for example so there are different fashion of alerting that they go to you but they go for school WMC right and if you're not doing any action you're going to have it you're gonna have someone show up with the describe to replace that display if you're still not doing any action we're gonna invoke how to might be proactive or sparing for that particular drive so it's a it's a continuous alert monitoring and alerting capability so we use this internally also for doing proactive support if the customer exceeds your expectations system do they care so will you just reporting or is there potential cost of both no we replace will please that's our we have yes software to reduce the way other guys that's how we ensure that they they'll pay for here in a monetized increase I mean it so I try to wear out a drive into the bindle lab multiple time I never be subsidy and what allows you know that when you finish the cycle of railroad wearing the way that the SSD fails is different from any media and the drive may be continuous for another two months you don't know that's why we both the proactive us pay at that point but here first you mentioned synchronous and asynchronous replication another answer did you just talk it through accidentally we went in too much detail on that so what options are there and also what integrations you have with things like srn yeah so we should pour totally SRM we always reported that so it's the base both of our integration with the VMware the other things is a async replication it's we use the same technology for Thailand block now we use a snap deep to do the replication and we connected to system we have a pure down to five minutes all right five minutes and now for CLI you can change that but at that point you should go with the mirror and we have the mirror capability so the mirror is really we acknowledged from both sides we have cache for insisting we expand our caching between the two so we make sure that everything comes in a to be cannot be acknowledged and we also going to introduce for file at the end of this year in the major solution for file it's a solution that we have on the NX so basically you can have a zero RPO for file without magic transparency over so the file one is really again we talked about the net service mer so we use the same concept of the net server string to take the internet server with the file systems with the snaps you're able to do a synchronous replication and because an ad server has the ideas with it when I fail over right I can take my business with its expanse print to your your host but I think I do need layer two if you say your failure is between the edge centers you have them I mean if you are Pyland you don't have layups you don't have access for your advice and so I'll give you dinner yes I will yeah you can get very specific uses not everyone would deployed specially in Europe there are it's typically in a metro their distances right like like two buildings or whatever within the same campus right at the very the most common things for a block is a sink and finally sink now we have the ability to send a snap that's another alternative that you can use so there is a different option you can have right now I think the most popular one is a secret location obviously but with this nap now also people they gonna start to use that and we said that now you can go to the card with a forever snap that's an awesome opportunity for you already see

Channel: Tech Field Day

Views: 8,328

Rating: 5 out of 5

Keywords: Tech Field Day, TFD, Storage Field Day, SFD, Storage Field Day 13, SFD13, Dell EMC, Dell EMC Unity, Wei Chen, CloudIQ, Unisphere

Id: MlFAeG4YlKQ

Length: 25min 25sec (1525 seconds)

Published: Mon Jun 19 2017