This week Nutanix announced the general availability release of our AOS™ 6.1 infrastructure software, delivering to customers higher performance, simpler management, enhanced orchestration capabilities, and more. The new functionality enhances the value of our customers’ HCI deployments and further simplifies deployment and provisioning, and customers can easily upgrade to this new version to benefit from these new capabilities.
An increasing number of IT organizations are standardizing on Nutanix® HCI to power the applications and services that run their businesses. We have been working closely with our customers to guide product development efforts and have simultaneously invested in our vision to lead the next era in IT by bringing powerful yet simple hybrid cloud infrastructure to organizations everywhere. Let’s take a closer look at the most impactful new functionality delivered with this release.
Host: Andy Whiteside
Co-host: Harvey Green
Co-host: Jirah Cox
This week Nutanix announced the general availability release of our AOS™ 6.1 infrastructure software, delivering to customers higher performance, simpler management, enhanced orchestration capabilities, and more. The new functionality enhances the value of our customers’ HCI deployments and further simplifies deployment and provisioning, and customers can easily upgrade to this new version to benefit from these new capabilities.
An increasing number of IT organizations are standardizing on Nutanix® HCI to power the applications and services that run their businesses. We have been working closely with our customers to guide product development efforts and have simultaneously invested in our vision to lead the next era in IT by bringing powerful yet simple hybrid cloud infrastructure to organizations everywhere. Let’s take a closer look at the most impactful new functionality delivered with this release.
Host: Andy Whiteside
Co-host: Harvey Green
Co-host: Jirah Cox
WEBVTT
1
00:00:03.210 --> 00:00:17.369
Andy Whiteside: hey everyone, welcome to episode 44 new tactics weekly i'm your host Andy whiteside as some of you guys know from listening we don't do a lot of like a spruce it up a lot, not a lot of intro music at exit music so Harvey decided, he would sing something for us.
2
00:00:19.980 --> 00:00:21.690
Harvey Green: twinkle twinkle little star.
3
00:00:24.420 --> 00:00:25.860
Andy Whiteside: Okay, good now twinkle twinkle little.
4
00:00:28.050 --> 00:00:33.030
Andy Whiteside: Alright sorry I put Harvey on the spot that i'd like to lighten the mood on a on a Monday.
5
00:00:34.260 --> 00:00:43.590
Andy Whiteside: harvey's with us Harvey green director of all kinds of things solutions architect as INTEGRA now director heads integrity of are you changing the world this week.
6
00:00:44.370 --> 00:00:46.200
Harvey Green: I am doing my best.
7
00:00:48.840 --> 00:00:50.310
Harvey Green: it's still early is still Monday.
8
00:00:51.000 --> 00:00:56.520
Andy Whiteside: Well, did I wrote he he has a whole new version of ios that he released he's doing his part oh.
9
00:00:56.580 --> 00:00:58.080
Jirah Cox: yeah I was very busy all weekend yeah.
10
00:00:58.830 --> 00:00:59.550
Jirah Cox: Finally, got around to it.
11
00:01:01.470 --> 00:01:11.640
Andy Whiteside: So that's that's the Voice of America job you guys have a new guy friend of our starting into tactics here in a few weeks, and he asked me what he needed and I was like just just know gyro just just make sure you know who gyrus that's all you need to know.
12
00:01:12.570 --> 00:01:15.510
Jirah Cox: Not not exactly scalable advice there but.
13
00:01:16.710 --> 00:01:19.140
Jirah Cox: I appreciate the compliment nonetheless yeah.
14
00:01:19.800 --> 00:01:24.450
Andy Whiteside: Well that's great very valuable resource lucky to have you on our podcast you.
15
00:01:24.510 --> 00:01:25.590
Jirah Cox: have to be here thanks for hosting.
16
00:01:26.370 --> 00:01:29.400
Andy Whiteside: So I said episode 44 I think they're wrong.
17
00:01:31.260 --> 00:01:35.880
Andy Whiteside: So episode 44 we're going to cover this blog from February 24 introducing.
18
00:01:37.320 --> 00:01:49.710
Andy Whiteside: Acropolis operating system acronym a os i'm a big fan of translating acronyms into their actual names, so that people get lost six dot one which is six period, one measure basically on them.
19
00:01:50.100 --> 00:01:50.490
Jirah Cox: I think we.
20
00:01:52.140 --> 00:01:53.010
Jirah Cox: are familiar with dots.
21
00:01:53.790 --> 00:01:54.240
Harvey Green: What is.
22
00:01:55.470 --> 00:01:56.820
Andy Whiteside: Why don't why don't we call this.
23
00:01:57.870 --> 00:02:02.730
Andy Whiteside: Acropolis operating system 2202.
24
00:02:05.880 --> 00:02:15.720
Andy Whiteside: You know, for the new naming Convention everybody's using you guys are still doing numbers and what point you hit this plateau and the numbers don't make sense anymore you're not doing Acropolis 99.8 like a radio station.
25
00:02:15.840 --> 00:02:23.580
Jirah Cox: Actually, the features here it's just not evenly distributed right actually PC is on the 2202 bandwagon so.
26
00:02:24.900 --> 00:02:29.190
Jirah Cox: I don't know, maybe I always get there one day, maybe it doesn't maybe maybe it stays a rebel.
27
00:02:30.510 --> 00:02:41.490
Andy Whiteside: Well, so i'm somewhat on topic and we're going to jump into HP here in a second, but your attendance is new tannic is full of acronyms give us your top 10 off top your head your own spot.
28
00:02:41.880 --> 00:02:47.370
Jirah Cox: Oh man, we ship like five new ones like last week alone we got like the ND bees and nc eyes and.
29
00:02:47.400 --> 00:02:48.120
Harvey Green: that's right.
30
00:02:48.210 --> 00:02:55.020
Jirah Cox: And us is and all kinds of good stuff yeah Of course you got all the oldies but goodies ALS is and CBS.
31
00:02:57.060 --> 00:03:01.890
Andy Whiteside: Whose fault is that is that the technical guys that like acronyms or is that the marketing guy.
32
00:03:03.090 --> 00:03:05.610
Andy Whiteside: Who does the acronym madness to ask, I can tell you.
33
00:03:05.850 --> 00:03:06.180
Jirah Cox: man.
34
00:03:06.570 --> 00:03:11.820
Andy Whiteside: i'm not a newbie, but I have not to newbies to come into them into his integrity and watching them just.
35
00:03:13.050 --> 00:03:16.560
Andy Whiteside: eyes roll back in their heads, all the acronyms I have no idea what any of means.
36
00:03:17.760 --> 00:03:19.980
Andy Whiteside: it's kind of a little bit of a problem in our industry.
37
00:03:21.060 --> 00:03:25.170
Jirah Cox: So it definitely is industry wide right the tlc is a three letter acronyms.
38
00:03:26.250 --> 00:03:26.610
Andy Whiteside: yeah.
39
00:03:28.260 --> 00:03:33.060
Andy Whiteside: Is it because people don't know what they stand for, is it because people, just like to be cool and use the acronyms.
40
00:03:33.180 --> 00:03:36.090
Jirah Cox: A tradition, mostly I guess.
41
00:03:36.540 --> 00:03:37.020
yeah.
42
00:03:38.040 --> 00:03:39.600
Andy Whiteside: we're just trying to shortcut anyway.
43
00:03:39.660 --> 00:03:42.540
Jirah Cox: I don't know i'm going I don't I don't get pulled in those meetings.
44
00:03:44.700 --> 00:03:56.610
Andy Whiteside: Well, so what we're talking about today is the new tannic celebrating system Acropolis operating system ios six dot one I think i've said that multiple times now it's it's new where was the previous release.
45
00:03:57.540 --> 00:04:05.520
Andy Whiteside: 606 dot oh so we're looking at an incremental change here, based on that nomenclature name nomenclature.
46
00:04:06.450 --> 00:04:07.410
Jirah Cox: yeah it's the.
47
00:04:10.170 --> 00:04:15.900
Jirah Cox: Second SDS short term support release in this trade yeah.
48
00:04:17.370 --> 00:04:24.210
Andy Whiteside: i'd love to know is, who do you think decides whether it's when it becomes a new whole number, the technical guys the marketing guys.
49
00:04:24.990 --> 00:04:25.740
i'm.
50
00:04:26.940 --> 00:04:29.220
Jirah Cox: Both I don't know more meetings that i'm not in.
51
00:04:29.700 --> 00:04:33.570
Andy Whiteside: yeah I don't know if you use a MAC or not use a MAC or windows.
52
00:04:33.840 --> 00:04:34.980
Jirah Cox: or use a couple of X yeah.
53
00:04:35.400 --> 00:04:36.360
Andy Whiteside: yeah no windows.
54
00:04:36.450 --> 00:04:38.430
Jirah Cox: I remote into windows computer.
55
00:04:38.790 --> 00:04:39.930
Andy Whiteside: yeah windows 10.
56
00:04:40.200 --> 00:04:42.390
Jirah Cox: or windows, the windows 10 yep yeah.
57
00:04:42.780 --> 00:04:50.430
Andy Whiteside: I mean that's a great example of no more operating system, who had seven new one coming out every month and then all sudden the marketing guys hey I think it's not renewing they're going here comes a new.
58
00:04:50.790 --> 00:04:54.540
Jirah Cox: yeah right and then somebody can go like oh it's one of those 11 verses you know.
59
00:04:55.500 --> 00:04:56.010
Jirah Cox: It was actually.
60
00:04:57.150 --> 00:05:01.830
Harvey Green: The last version of windows you'll ever have and you'll ever need never be another one.
61
00:05:02.700 --> 00:05:03.390
Harvey Green: You know the bully.
62
00:05:03.750 --> 00:05:05.700
Jirah Cox: Both keeping up with each other, now they're both on 11.
63
00:05:06.600 --> 00:05:08.400
Andy Whiteside: i'm not gonna lie, I thought this was gonna be the top.
64
00:05:09.120 --> 00:05:09.930
Harvey Green: Oh.
65
00:05:10.830 --> 00:05:12.120
Jirah Cox: This is look at solder.
66
00:05:13.500 --> 00:05:17.550
Jirah Cox: People to inform us to hear our commentary about what a 15 year old os.
67
00:05:19.200 --> 00:05:26.730
Jirah Cox: probably never actually ran as their daily driver, but no vista definitely got got the short end of the stick there like it was unfairly maligned no.
68
00:05:27.300 --> 00:05:27.930
well.
69
00:05:29.280 --> 00:05:30.900
Andy Whiteside: windows me might have got it worse.
70
00:05:31.920 --> 00:05:33.150
Jirah Cox: windows, I mean deserved it.
71
00:05:38.790 --> 00:05:44.640
Andy Whiteside: Alright, so a crop this operating system six dot one incremental release gyro.
72
00:05:45.870 --> 00:05:56.310
Andy Whiteside: what's the does the intro in this blog have anything of relevance to it that kind of tips its hat to why six dot one came about, are we going to find that here in a minute we start going through it.
73
00:05:57.510 --> 00:05:59.910
Jirah Cox: I think it's pretty pretty awesome pretty well rounded release.
74
00:06:01.590 --> 00:06:09.630
Jirah Cox: Making yeah i'll say i'll say just agree with the intro right making on Prem and hybrid clouds everywhere just better overall.
75
00:06:11.730 --> 00:06:14.670
Andy Whiteside: yeah like oh see, I have a new guy that's going to be working for you guys I believe and.
76
00:06:15.450 --> 00:06:23.610
Andy Whiteside: He was a trying to understand from the interview process is new tannic software for the cloud or on premises, because it sounds like you just cloud.
77
00:06:23.970 --> 00:06:35.880
Andy Whiteside: And you know, obviously that's where your company wants people to go and they want to go with you at the same time, you know it's one heck of a cloud software that enables on premises and transitions into the cloud or or vice versa right back and forth.
78
00:06:36.210 --> 00:06:37.860
Jirah Cox: it's software for wherever your cloud is.
79
00:06:38.370 --> 00:06:42.690
Andy Whiteside: that's right defined cloud or better yet defined clouds and we can have that conversation.
80
00:06:44.880 --> 00:06:55.140
Andy Whiteside: Alright, so section one talks about the memory, or the first non introductory section memory overcommit for HIV, which stands for Acropolis hypervisor.
81
00:06:56.820 --> 00:07:01.890
Andy Whiteside: I tell me where that didn't work before where that was sort before and what you guys have done now.
82
00:07:02.700 --> 00:07:16.230
Jirah Cox: Sure yeah basically before it was all memory was one to one assigned blocked right So if you promised a vm eight gig memory, then we had to go find it a host somewhere and allocated for just that vm use, so it wasn't over committed wasn't shared.
83
00:07:17.460 --> 00:07:25.680
Jirah Cox: With this now, you can promise more memory to your virtual machines than they actually have present on the hosts in the cluster right so.
84
00:07:26.040 --> 00:07:30.600
Jirah Cox: something you know to do on other other hypervisor is for a while we're bringing it down to HP as well.
85
00:07:31.260 --> 00:07:43.860
Jirah Cox: With us exact same design considerations that have always existed elsewhere have if you promise too much and then use too much we can't make more memory for free that's a hardware defined limit, not a software defined limit.
86
00:07:44.910 --> 00:07:56.550
Jirah Cox: So use with us with consideration us carefully but it's fantastic for like live environments, you know development environments where you know you need to be able to spin up you know hundreds of vm.
87
00:07:58.560 --> 00:08:04.260
Jirah Cox: And, and the usage windows vm is very dynamic I had one customer years ago, who.
88
00:08:04.800 --> 00:08:16.320
Jirah Cox: Who every every night at midnight just powered off all other the other delphiniums and if you didn't come into work, the next day if you were off that day, then you your vm didn't take up any resources right you just powered them on when you got your coffee in the morning.
89
00:08:17.670 --> 00:08:30.030
Jirah Cox: So those kind of environments right where you aren't particularly picky about performance for any one single vm you just want you know, to let everybody see what needs to be seen by the at the os level absolutely great feature.
90
00:08:30.420 --> 00:08:43.470
Harvey Green: So let me highlight something that you said there and and do that by highlighting something you didn't say there you said Dev and you said test, but you did not say pride why.
91
00:08:45.060 --> 00:08:49.080
Jirah Cox: it's a it's totally up to the customer right it's what are your requirements for your environment.
92
00:08:49.590 --> 00:09:03.330
Jirah Cox: Personally, you know what I coach my customers is you know something to turn on for prod day one, maybe not right, which is why we've we've taken a very cautious approach to even if developing it shopping at HP is what seven seven years old now eight years old now.
93
00:09:05.010 --> 00:09:15.120
Jirah Cox: Because the as soon as you overcommit memory, then you have an opportunity for overdemand of memory and you have to get an interesting fun things like paging.
94
00:09:16.320 --> 00:09:23.100
Jirah Cox: and other other fun behaviors right that guess what are just not even something have to worry about with our memory overcommit enabled.
95
00:09:23.490 --> 00:09:25.350
Harvey Green: So i'll have to tell you this.
96
00:09:27.120 --> 00:09:44.280
Harvey Green: One of the, I guess, I will call it a safety one of the safeties that I always was happy that Nintendo his head was this one, where you could not just go out there and over commit your memory by you know 221321.
97
00:09:46.050 --> 00:09:51.750
Harvey Green: And so I like that this is in there and I also don't like this is.
98
00:09:53.460 --> 00:09:59.250
Jirah Cox: i'm with you i'm totally with the rv and and that's why it's a customer by customer environment by environment choice.
99
00:10:00.150 --> 00:10:05.070
Jirah Cox: to opt in i'm with you, if it was my environment to run I probably wouldn't enable it for production.
100
00:10:06.030 --> 00:10:10.560
Jirah Cox: i'm it took a meeting with one customer years ago, who they said, you know, without.
101
00:10:11.130 --> 00:10:20.400
Jirah Cox: Without this feature you'll never you know go and come in our data Center right in a major way, at least, and we looked at their V Center and they had they were running for two one in production right.
102
00:10:20.820 --> 00:10:41.640
Jirah Cox: At the at the hundreds of terabytes of memory scale and i'm like Okay, I can see your point where, if you're running this and managing it safely today, then, if I say you need to buy for X as much memory to run hv that might be challenging so you know use what you use wisely yes.
103
00:10:42.000 --> 00:10:47.700
Andy Whiteside: So, so I had forgotten this was in here because we never tell anybody that overcommit, even though I do understand.
104
00:10:48.570 --> 00:11:01.350
Andy Whiteside: That it makes me makes some sense to some people, the two questions for you guys one, what do they do in hyper scale lawyers, do they overcommit you and you don't realize it or you opt in or opt out what's How does it work there.
105
00:11:03.600 --> 00:11:04.110
Andy Whiteside: Do we know.
106
00:11:04.410 --> 00:11:06.480
Jirah Cox: I I think I couldn't give you a firm answer.
107
00:11:07.710 --> 00:11:11.250
Jirah Cox: They probably would tell you they probably would say doesn't matter if you can't tell.
108
00:11:11.880 --> 00:11:16.080
Jirah Cox: yeah right managing of what overcommit ratio might exist.
109
00:11:16.950 --> 00:11:24.270
Andy Whiteside: So here's the second part of that question, the second question, maybe they're tied together Harvey if you Okay, so if you don't do it in memory.
110
00:11:25.980 --> 00:11:33.420
Andy Whiteside: In other words, if you overcommit and it's not able to happen in real physical memory and it has to go somewhere else, where does it go.
111
00:11:36.210 --> 00:11:38.310
Harvey Green: me make sure I understand the question if.
112
00:11:38.820 --> 00:11:39.930
Jirah Cox: it's a question or pop quiz.
113
00:11:42.720 --> 00:11:50.370
Andy Whiteside: So if you if you overcome it and you run out of physical memory and it has to start paging or you know traditional paging.
114
00:11:51.450 --> 00:12:00.450
Andy Whiteside: So, I guess, a new machine if it runs out in this case it thinks it still has it does it does the system start doing paging page swaps for you.
115
00:12:00.930 --> 00:12:08.940
Jirah Cox: Right yeah usually that's same thing happens inside the ios can happen outside the ios inside the osu page into your windows page file or Linux page.
116
00:12:10.110 --> 00:12:18.120
Jirah Cox: Like swap partition and, similarly, a vm can page out to you know, have the EMS allocated page files as well, but.
117
00:12:18.270 --> 00:12:21.480
Andy Whiteside: If i'm doing that in a system where i'm going to a file and i've got.
118
00:12:22.740 --> 00:12:28.620
Andy Whiteside: You know envy me or ssd that slow the slow i'm i'm still pretty darn fast right.
119
00:12:30.510 --> 00:12:30.750
Jirah Cox: or.
120
00:12:32.970 --> 00:12:34.560
Jirah Cox: Perhaps, maybe for a certain point of view.
121
00:12:34.560 --> 00:12:34.950
Although.
122
00:12:35.970 --> 00:12:45.390
Jirah Cox: I probably would agree with that right if it's if a vm things its memory unexpected memory speed that your orders of magnitude slower, even going to disk there.
123
00:12:46.500 --> 00:13:00.780
Harvey Green: Well, and so the other thing that I would bring up is even if I ignore the rest of what's going on the fact that you are now paging potentially to a physical disk at this point.
124
00:13:01.260 --> 00:13:13.770
Harvey Green: makes you nowhere near as portable as you used to be, and so, in the event of a disk failure or node failure, all of the all of the magic of new tactics and this you know.
125
00:13:14.580 --> 00:13:29.250
Harvey Green: resilient system is no longer in my eyes as resistant and me as resilience and maybe they've got some more magic on the back end that I just don't know about yet, but I just don't like the setup.
126
00:13:30.630 --> 00:13:35.010
Jirah Cox: Well, and also lastly there's probably no world where that makes economic sense if you.
127
00:13:35.130 --> 00:13:37.050
Jirah Cox: want X amount of memory.
128
00:13:37.080 --> 00:13:44.430
Jirah Cox: Sorry that's a harsh answer but probably true if you bought X amount of memory and then you fill it all up and you clearly objectively need more.
129
00:13:45.810 --> 00:13:49.290
Jirah Cox: The storage tier there isn't a costume more than buying a memory that you clearly made.
130
00:13:49.530 --> 00:13:56.640
Jirah Cox: Right so maybe it's a stopgap emergency method, but but, in reality, like you just buy more memory by what you need yeah.
131
00:13:57.420 --> 00:14:02.430
Harvey Green: This is why I both like that it's there and I don't like that it's the.
132
00:14:03.570 --> 00:14:11.040
Jirah Cox: Time in a place right that what I ironically given you know our venn diagram here of our skill sets I can picture.
133
00:14:12.150 --> 00:14:17.880
Jirah Cox: Any you see environment being actually punish a great candidate right very, very thirsty workloads.
134
00:14:19.230 --> 00:14:32.310
Jirah Cox: Lack of lack of very thirsty a lot of not a lot of insight into like potentially what's my our by our forecasted need right, so I can make some promises here for the environment in general paint with a broad brush and then trust the ball sort of self out.
135
00:14:34.110 --> 00:14:51.480
Andy Whiteside: Well, so here's what I was getting out with all that I get the way we used to be scared of doing this when we had slower drives we had less smart architecture, but with the Acropolis operating system, control and storage layer i'm less scared of overcoming, then I was back then.
136
00:14:53.220 --> 00:15:01.200
Harvey Green: with which I agree with the merits of that and that's honestly why it makes me even more risky because.
137
00:15:02.490 --> 00:15:23.700
Harvey Green: I could see the potential of somebody you know, taking that mindset and then over committing on purpose on to the point that they can, to the point that, when they do actually need to overcome if they don't have that as a luxury anymore.
138
00:15:24.270 --> 00:15:27.900
Andy Whiteside: Well that's why I brought up, I wonder what the hyper steelers do they don't tell us do they they.
139
00:15:28.260 --> 00:15:31.590
Andy Whiteside: Because they're Point two divers like you can't tell, then you don't need to know.
140
00:15:33.750 --> 00:15:34.830
Harvey Green: that's exactly right.
141
00:15:36.300 --> 00:15:44.820
Harvey Green: And maybe maybe this is the same way I haven't seen it in action, yet, and maybe maybe i'm being overly cautious but.
142
00:15:46.170 --> 00:15:47.970
Harvey Green: i'll be overly cautious.
143
00:15:48.270 --> 00:15:56.340
Andy Whiteside: It would be neat to be able to go to the hyper scales and figure it out that way for examples INTEGRA has new tactics running in our data centers we offer that as a cloud, we could.
144
00:15:56.610 --> 00:16:07.830
Andy Whiteside: We could say that if you check this box or sign on this agreement you've never had that happen you don't have to come check behind us to try and see if you're getting over committed or not, it just you know here's an agreement they won't happen.
145
00:16:09.330 --> 00:16:09.600
yeah.
146
00:16:11.940 --> 00:16:18.150
Andy Whiteside: All right, let's go to the topic we all love talking about the most which is databases how we're going to make these things scale up.
147
00:16:19.260 --> 00:16:22.170
Andy Whiteside: In the new Acropolis operating system six that one.
148
00:16:23.640 --> 00:16:25.800
Jirah Cox: yeah, so this is, this is a very exciting.
149
00:16:26.880 --> 00:16:34.140
Jirah Cox: enhancement that, in a nutshell right we call it kind of vetoes charting but basically it lets a virtual disk controller.
150
00:16:35.400 --> 00:16:39.990
Jirah Cox: run in a multi threaded fashion, so we can get multiple paths to get to the same disk.
151
00:16:41.730 --> 00:16:49.110
Jirah Cox: That dramatically in this in this release speeds up the Read performance is quite a bit, so this you know we've said, for years, we still say.
152
00:16:49.650 --> 00:16:58.080
Jirah Cox: The best way to get performance out of your databases with multiple disks it's also in like the sequel and Oracle best practices right lots of disks and spread your data files across them.
153
00:16:59.250 --> 00:17:07.860
Jirah Cox: First, you know, some people when they first moved to new tannic then that's maybe not practical they can't get the downtime to take the database offline and relay it out onto multiple data files.
154
00:17:08.460 --> 00:17:15.510
Jirah Cox: So we still want to ship enhancements that helped them out where they are right so help those single large databases run faster.
155
00:17:16.170 --> 00:17:29.310
Jirah Cox: And so that's what this enhancement is doing here is improving the performance for that by some some really dialing numbers you don't we don't want to you know read bar charts here into audio form but take a look at it it's it's very compelling stuff.
156
00:17:31.200 --> 00:17:42.480
Andy Whiteside: We could use the numbers, the percentage number, but let me just visual explain the visual here in terms of what it does so we've got our hypervisor inside the hypervisor is a control vm star controller.
157
00:17:43.500 --> 00:17:49.680
Andy Whiteside: It runs stargate real quick direct tin tin 22nd definition what stargate is.
158
00:17:49.920 --> 00:18:02.640
Jirah Cox: Yes, very good is the process by which you know, a hypervisor you know sends io requests to mechanics and then we figure out what node and what disk needs to service all of those stargate because you can kind of go in anywhere and come out anywhere else.
159
00:18:03.780 --> 00:18:11.370
Andy Whiteside: And then that hands off to the V disk controller we talking to software inside the control vm I guess to be are based on the.
160
00:18:11.400 --> 00:18:16.170
Jirah Cox: Right so like what what part of all what part of our code owns like your C drive or your D drive.
161
00:18:16.740 --> 00:18:17.070
yeah.
162
00:18:18.690 --> 00:18:22.680
Andy Whiteside: And then now we're getting down to a single V disk in parallel.
163
00:18:23.730 --> 00:18:25.740
Andy Whiteside: reads and writes or just reads.
164
00:18:26.070 --> 00:18:28.350
Jirah Cox: This one enhances rates OK.
165
00:18:29.820 --> 00:18:30.120
Andy Whiteside: OK.
166
00:18:33.150 --> 00:18:40.680
Andy Whiteside: So now, I mean i'm playing devil's advocate through well okay so it's always important if i'm the the storage guy I want to.
167
00:18:40.950 --> 00:18:54.150
Andy Whiteside: Okay, I want my readers to be right, but I sure as heck one my rights to be right, so what you guys are doing here is you're making the reads a lot faster maybe a step one to doing this process getting more getting more performance out of access to that data.
168
00:18:54.870 --> 00:19:00.660
Jirah Cox: Right and number of reads and writes can follow different code paths so that's why you can see an enhancement that hits one of those and not the other yeah.
169
00:19:01.980 --> 00:19:13.290
Andy Whiteside: And then, if I go down here to the bar chart i've got the green bar and now the blue bar and it looks like on the right hand side i'm getting significantly more something's.
170
00:19:13.980 --> 00:19:16.080
Jirah Cox: that's on the left there right the use case you know.
171
00:19:16.290 --> 00:19:25.200
Jirah Cox: single V disk right, so if you've moved from like a three tier deployment of your database and moved on new tonics without sort of following our best practices.
172
00:19:25.800 --> 00:19:29.040
Jirah Cox: sequel Oracle best practices still on a single large V disk.
173
00:19:29.520 --> 00:19:43.860
Jirah Cox: Performance improvement there, but then, also on the right side if you are deploying it cornea best practices, even more benefit for you there right, because now if you're deployed across 888 disk and the system those all benefit from the improvement as well, so, then you get even more.
174
00:19:44.880 --> 00:19:46.470
Jirah Cox: even faster read enhancement.
175
00:19:49.980 --> 00:20:00.000
Andy Whiteside: Well, and that just continues to show that new tannic serious about getting those naysayers on the database side over to new tactics.
176
00:20:00.120 --> 00:20:00.480
yeah.
177
00:20:04.230 --> 00:20:08.730
Andy Whiteside: um let's move on to the next one, because I know Harvey didn't want to talk about the database, one that much.
178
00:20:12.000 --> 00:20:18.450
Andy Whiteside: You have a you have any dog in that database file Harvey are you just gonna say yeah I did uh.
179
00:20:20.790 --> 00:20:22.140
Harvey Green: yeah do.
180
00:20:24.450 --> 00:20:31.020
Andy Whiteside: The next one talks about something that sounds extremely interesting octane slash envy me.
181
00:20:31.620 --> 00:20:47.610
Andy Whiteside: tearing I know envy me has been a hot topic at least for me the last couple years because anything fast and anything fastest good anything unbelievably fast is good he's a better it's over what what are you guys doing it this piece of the enhancements.
182
00:20:47.850 --> 00:20:53.460
Jirah Cox: yeah I think I think this first graphic kind of tells a really fun story where it shows you know hard disk drives spinning, you know magnetic.
183
00:20:53.880 --> 00:21:05.640
Jirah Cox: Storage less than 500 I OPS per device and ssd maybe call it 20,000 I OPS per device envy me ssd 400,000 and obtain maybe 550,000 so.
184
00:21:05.970 --> 00:21:20.910
Jirah Cox: All still crazy fast, but there's a little bit of daylight between the envy me performance, the paint performance with this enhancement now as storage can differentiate between those two things so used to be, they were both just so crazy fast, they would go into one tier together.
185
00:21:22.350 --> 00:21:28.710
Jirah Cox: The the tearing them out here now lets us, you know deliver say like obtain tier performance.
186
00:21:29.610 --> 00:21:38.010
Jirah Cox: To the very hottest subsets of data right like think back to when we you know our first launched with our very tiny as as DS and spinning disks.
187
00:21:38.520 --> 00:21:47.940
Jirah Cox: Are hybrid nodes right but, but you can get very, very fast performance out of them, because all the hot data went on to the hottest tier of storage on in that case ssd.
188
00:21:48.390 --> 00:21:57.750
Jirah Cox: This brings back that same kind of functionality, right now, if I had the system that had octane plus envy me drives my very hardest data goes into the very hottest tier of storage.
189
00:21:59.370 --> 00:22:08.010
Andy Whiteside: So spinning hard drives were great back in the day, and then we got to ssd and then we got to India me now the fastest thing is Intel obtain.
190
00:22:08.340 --> 00:22:15.510
Jirah Cox: yep yeah it's a it's it's candidly very weird to say you can configure system where envy me is the slow tier.
191
00:22:17.400 --> 00:22:21.780
Andy Whiteside: Where does the Intel obtain live in regard to the system board.
192
00:22:23.490 --> 00:22:30.570
Jirah Cox: It looks like a regular desk or you'd see it next year, you know ssd it's still another form of solid state, but that works kind of diluted right.
193
00:22:32.070 --> 00:22:34.380
Jirah Cox: it's another form of stories that we just live in a disk slot.
194
00:22:35.520 --> 00:22:53.820
Andy Whiteside: And I was under the impression end me was partially fast because of where it's set in relation to the system board and not having to go through a storage controller is the Intel obtain just so fast going through a controller is irrelevant and it's just reads and writes fs.
195
00:22:54.000 --> 00:23:03.780
Jirah Cox: My understanding is that both of them are are you know pci devices right so and both of them can bypass control or they just become become pci E devices.
196
00:23:08.730 --> 00:23:09.240
Andy Whiteside: yeah I mean.
197
00:23:10.290 --> 00:23:11.070
Andy Whiteside: What needs that.
198
00:23:12.600 --> 00:23:21.720
Harvey Green: yeah i'm sitting here, just like shaking my head, because all I hear is it went from crazy fast to stupid fast.
199
00:23:23.790 --> 00:23:30.120
Jirah Cox: And that's part of the value prop right, we want the software to be ready for the hardware and the demands of tomorrow.
200
00:23:31.500 --> 00:23:38.310
Jirah Cox: And and evolve to let you the customer take advantage of the latest and greatest hardware without changing your management.
201
00:23:38.970 --> 00:23:50.340
Jirah Cox: plane without having to rework the workloads, you know as the hardware gets better it's easy to add those kind of nodes to the cluster move workloads on to it inject the oldest nodes in the cluster and get on with your life.
202
00:23:51.030 --> 00:24:11.070
Harvey Green: yeah i'm sitting here, looking at this chart and it's talking about the latency between the two and starts out on the left with hard drives it more than two milliseconds then we get into microseconds and here we are going from less than 100 microseconds all the way down to 10 so.
203
00:24:12.810 --> 00:24:13.650
Harvey Green: This is.
204
00:24:15.060 --> 00:24:17.610
Harvey Green: whoo that's that's fast.
205
00:24:18.630 --> 00:24:18.990
Jirah Cox: It is.
206
00:24:19.140 --> 00:24:21.330
Harvey Green: very fast so.
207
00:24:21.360 --> 00:24:25.800
Andy Whiteside: dire we have a nice diagram kind of a flow workflow here on the screen wants you to explain this.
208
00:24:30.630 --> 00:24:33.960
Jirah Cox: Is a verbal whiteboard in a in an audio medium.
209
00:24:34.320 --> 00:24:36.840
Jirah Cox: And it shows the right path right now.
210
00:24:37.860 --> 00:24:46.800
Jirah Cox: reads can come from lots of places right both persistent storage and you know whether that's hot or cold to your storage, as well as even.
211
00:24:47.280 --> 00:24:54.660
Jirah Cox: The cash that can live within the CRM itself and then rights go through our blog and then find their way to where they need to live elsewhere in the system as well.
212
00:24:55.680 --> 00:24:55.920
yeah.
213
00:25:01.500 --> 00:25:22.470
Andy Whiteside: This next there's another bar chart here it basically just talks about either all envy me versus envy me plus obtain so help me understand this in the case of envy me plus obtain, would you be using which one would be the fast one octane, of course, right.
214
00:25:23.820 --> 00:25:32.100
Jirah Cox: Right yeah yeah and all envy me versus and we may pose an opt in to your updating devices are like tiny right think about like 200 gig 400 gig.
215
00:25:33.540 --> 00:25:42.330
Jirah Cox: If I can get a bigger than that but, like commonly deployed sizes are there they're tiny but that latency is crazy low right so letting the hottest amount of data flow on to those.
216
00:25:44.250 --> 00:25:50.310
Andy Whiteside: lungs obtain that around because I I thought I was the cool kid talking about envy me and I didn't know, I was dating myself.
217
00:25:50.610 --> 00:25:51.480
um.
218
00:25:52.860 --> 00:25:57.150
Jirah Cox: I mean Wikipedia probably would say like five six years, maybe what venture to guess on my part, but.
219
00:25:57.990 --> 00:25:59.370
Harvey Green: Just a couple years.
220
00:25:59.430 --> 00:26:02.700
Jirah Cox: takes a while to find stuff for free protect find its way into the data Center right.
221
00:26:02.910 --> 00:26:12.510
Andy Whiteside: Right right alright so here's a question and obtain drive computer didn't VOC a terabyte of August, you just said what's the biggest opportunity would find Dr.
222
00:26:12.990 --> 00:26:18.510
Jirah Cox: I the ones I feel like I handle commonly in a configuration or yeah like 200 gig 400 gig.
223
00:26:19.230 --> 00:26:25.410
Andy Whiteside: So 200 gig versus 200 gig India me versus 200 gig ssd if there is such thing anymore.
224
00:26:25.710 --> 00:26:27.810
Andy Whiteside: But what give me pricing like.
225
00:26:27.870 --> 00:26:29.610
Jirah Cox: I are not a hardware guy.
226
00:26:31.290 --> 00:26:32.850
Jirah Cox: I mean newegg.com I don't know.
227
00:26:34.890 --> 00:26:35.190
Andy Whiteside: Okay.
228
00:26:36.450 --> 00:26:37.500
Andy Whiteside: All right, i'm learning a lot here.
229
00:26:38.550 --> 00:26:52.260
Andy Whiteside: Alright next section talks about the enhanced business continuity encrypted Dr traffic improve ios vss provider let's go with the encrypted Dr traffic that's pretty self explanatory but.
230
00:26:52.650 --> 00:26:58.470
Jirah Cox: yeah maybe some of harvey's customers over on on the on the Gulf side you know have stronger preferences around.
231
00:26:58.830 --> 00:27:10.440
Jirah Cox: I want all the application traffic from A to B to be encrypted and now that's an easy that's an easy, yes, without needing to bring in like another vpn tunnel or doing like switch to switch IP SEC.
232
00:27:13.560 --> 00:27:14.520
Andy Whiteside: yep so easy enough.
233
00:27:15.570 --> 00:27:22.080
Andy Whiteside: Improved ALS so across the substrate operating system vss provider.
234
00:27:23.610 --> 00:27:26.130
Andy Whiteside: I think I should know what vss means but somebody educate me.
235
00:27:27.150 --> 00:27:39.840
Jirah Cox: that's the volume snapshot service, I might have that a little bit off, but that's because what it does, is it lets you it's it's how you can do in gas gas right for in guest data when you're taking a backup.
236
00:27:42.540 --> 00:27:46.410
Jirah Cox: commonly of stuff like sequel or outlook back in the day when we ran outlook on Prem.
237
00:27:48.510 --> 00:27:58.680
Jirah Cox: So with these enhancements you can do stuff like incremental backups or, if you have multiple vss writers, you can target which one you want to have class for the given snapshot.
238
00:28:00.780 --> 00:28:03.840
Andy Whiteside: I just I used to shut on my vm down take my backups.
239
00:28:05.700 --> 00:28:07.410
Andy Whiteside: there's not there's not a secret anymore I don't.
240
00:28:07.890 --> 00:28:10.170
Jirah Cox: throw the tapes in the trunk drive to the safe deposit box.
241
00:28:12.270 --> 00:28:16.050
Andy Whiteside: I couldn't take my brakes until the storage that tape backup guy showed up every day.
242
00:28:20.550 --> 00:28:23.490
Andy Whiteside: Okay Harvey any comments on either one of those.
243
00:28:24.540 --> 00:28:28.530
Harvey Green: No, I mean this, these are all very, very good things I like.
244
00:28:30.810 --> 00:28:31.680
Jirah Cox: harvey's a fan.
245
00:28:32.700 --> 00:28:44.370
Andy Whiteside: Alright next section talks about deeper storage inside starts with a E s on hybrid clusters, or they were at least are nice enough to give us the what he stands for autonomous extent store.
246
00:28:45.960 --> 00:28:48.240
Andy Whiteside: What is this on hybrid clusters do for me.
247
00:28:48.720 --> 00:28:50.250
Jirah Cox: So, as we shipped.
248
00:28:51.510 --> 00:28:55.110
Jirah Cox: miss a couple years ago in round terms, it was one of our next.
249
00:28:56.370 --> 00:28:58.620
Jirah Cox: headlines that brought.
250
00:29:00.480 --> 00:29:10.230
Jirah Cox: yeah this feature called upon a success story, which I describe it when i'm whiteboarding for customer has made a data locality right, so you you write data in your vm.
251
00:29:10.710 --> 00:29:26.040
Jirah Cox: We write metadata as the storage provider around where that data lives in the system right So if you wrote a megabyte of data that our system is to record what node is storing that and on what desk and then what other node and what other disk system is storing it for like say rf two.
252
00:29:27.540 --> 00:29:35.190
Jirah Cox: minutes to locality says hey with it with a is here, what if we let the cluster not worry about all the details right.
253
00:29:35.700 --> 00:29:46.770
Jirah Cox: If the node wants to move from disk one to disk disk to that's a node local decision and the rest of us are doesn't have to care about that right so less Meta data stored globally.
254
00:29:47.670 --> 00:29:59.280
Jirah Cox: mean smaller databases means faster performance and then, so what this is really talking about here is now that functionality comes to hybrid notes so yea it was all flash and faster nodes for a long time.
255
00:30:00.240 --> 00:30:04.710
Jirah Cox: So now even even your ssd plus spinning this systems get to get to benefit from that.
256
00:30:05.850 --> 00:30:06.930
Jirah Cox: enhance performance as well.
257
00:30:10.080 --> 00:30:11.400
Andy Whiteside: or read comments.
258
00:30:13.230 --> 00:30:14.010
Harvey Green: nope.
259
00:30:14.310 --> 00:30:15.420
Harvey Green: harvey's a fam like it.
260
00:30:17.430 --> 00:30:19.920
Harvey Green: These these are like checking boxes for me right now.
261
00:30:22.080 --> 00:30:26.730
Andy Whiteside: Alright next section talks about storage consumption reporting improvements.
262
00:30:27.930 --> 00:30:32.070
Andy Whiteside: Probably pretty straightforward but jarrod give us give us what this does and what it means for.
263
00:30:33.120 --> 00:30:33.450
Jirah Cox: yeah.
264
00:30:34.530 --> 00:30:38.280
Jirah Cox: Totally beefing up the storage widget there on the prism dashboard.
265
00:30:39.900 --> 00:30:50.940
Jirah Cox: biggest one I love here is showing the snapshot usage versus actual data storage usage right so more details more at your fingertips easier to use good stuff.
266
00:30:54.150 --> 00:30:59.280
Andy Whiteside: Have you can you think of a time where this information would have been valuable that you didn't have.
267
00:30:59.610 --> 00:31:00.360
Yes.
268
00:31:02.340 --> 00:31:21.300
Harvey Green: Literally exactly what you just described, where a customer was storing a bunch of snapshots and did not did not fully recognize why they were lacking in storage and you know, sometimes.
269
00:31:22.560 --> 00:31:30.600
Harvey Green: From a customer standpoint, you want to make sure you have I mean try this customer they they wanted to have a lot of.
270
00:31:32.190 --> 00:31:40.440
Harvey Green: A lot of backup copies because they had just emerged from a ransomware event that they did not discover for a couple weeks.
271
00:31:41.010 --> 00:31:47.550
Harvey Green: And so they you know, one of the ways that they wanted to move forward was to make sure that they had.
272
00:31:48.000 --> 00:31:58.320
Harvey Green: A month worth of snapshots to go back to, but there was no real thought given to what kind of space that would take up for them.
273
00:31:59.070 --> 00:32:19.620
Harvey Green: So they started eating through a lot of space and wondering why in the world they're fairly new system was not giving them the amount of storage space that they expected and being able to see this much more easily would have saved them at least some time and heartache.
274
00:32:20.640 --> 00:32:36.840
Jirah Cox: yeah I can think of that or you know it's it's more fashionable nowadays to have like say two data centers and you don't call them prod and Dr right we're one is all running all the all the workloads one's running none.
275
00:32:37.950 --> 00:32:51.930
Jirah Cox: Now you're typically more active active and you run half the things on each one, and of course it's it's still the case that if you can do application level availability that's always better than infrastructure level availability like active directory does application level availability.
276
00:32:55.890 --> 00:33:01.830
Jirah Cox: With that right have had a cluster that was running some vm but also receiving replicas of other vm from other sites.
277
00:33:02.310 --> 00:33:16.410
Jirah Cox: This would be immediately useful to say here's what the clusters consuming locally for its own vm maybe it's replicating those elsewhere and then other vm or our incoming by replication what's the space breakdown between those two kinds of buckets of data, yes.
278
00:33:17.070 --> 00:33:24.270
Andy Whiteside: Is it fair to say this data was available, but at a lower level somewhere in the Shell somewhere versus in the ui.
279
00:33:24.510 --> 00:33:29.820
Jirah Cox: yep totally agree more up to the surface surface and easily reportable.
280
00:33:30.930 --> 00:33:39.690
Andy Whiteside: I mean it may or may not have happened but way back in the day I might have been able to find this stuff through a cry, but I just started deleting snapshots until I saw the desired result that might have happened I don't.
281
00:33:41.280 --> 00:33:45.930
Jirah Cox: mean between that and and Andy just letting them rip with.
282
00:33:47.070 --> 00:33:50.550
Jirah Cox: memory over consumption Harvey i'm hearing you need to double check so.
283
00:33:53.160 --> 00:33:55.620
Andy Whiteside: They haven't let me out of the lab scenario and a long.
284
00:33:55.620 --> 00:33:57.090
Andy Whiteside: Time oh okay.
285
00:33:59.760 --> 00:34:09.510
Andy Whiteside: All right, enhanced virtual networking, the first thing I saw when I glanced down was this three letter acronym vpn what's what's this thing about the southern that extension piece.
286
00:34:10.320 --> 00:34:17.520
Jirah Cox: So vpn or oh yeah right and then also really vpc as well, so like virtual virtual private clouds.
287
00:34:18.720 --> 00:34:28.080
Jirah Cox: So with this now with the flow networking, you can extend your sub nets across sites now, which is very exciting so that.
288
00:34:28.470 --> 00:34:38.700
Jirah Cox: That you can now have a software to find something that that you know you see it as spanning both of these two data centers in reality that you have no network harbor supporting that and right all done in software.
289
00:34:40.290 --> 00:34:47.190
Andy Whiteside: yeah I could see this being very values, is there any type of latency requirement sub sub millisecond to do this, or is it doesn't matter.
290
00:34:48.930 --> 00:34:57.270
Jirah Cox: I don't have a chapter in verse memorized on that one I assume laws of physics still apply So if you have bad latency you can probably not have a great day, turning this on.
291
00:34:58.140 --> 00:35:04.290
Andy Whiteside: yeah but you talked a while ago about having you know active active data centers workloads in both places.
292
00:35:04.320 --> 00:35:04.740
Jirah Cox: Totally.
293
00:35:05.010 --> 00:35:11.820
Andy Whiteside: I guess you could have application performance issues but, in general, your data Center just logically became two places.
294
00:35:12.240 --> 00:35:21.300
Jirah Cox: 100% yeah give me good software defined networking is built on a foundation of good hardware defined networking right so thou shalt have good latency and throughput.
295
00:35:21.750 --> 00:35:29.670
Jirah Cox: And then you have a great experience, using a software defined overlay On top of that, for ease of management and mobility which.
296
00:35:30.150 --> 00:35:32.370
Harvey Green: Which direct commandment laws that.
297
00:35:33.240 --> 00:35:34.170
Harvey Green: have good latency.
298
00:35:34.890 --> 00:35:37.050
Jirah Cox: We call it number one have a good network.
299
00:35:39.330 --> 00:35:52.440
Andy Whiteside: It was funny I was in Cancun, last week I couldn't communicate with anybody I tried every APP I can think of friend of mine recommended I use, you know what's that he's like oh my European friends yeah like yeah but if I have bad networking it doesn't matter whatever i'm using.
300
00:35:54.360 --> 00:36:01.050
Andy Whiteside: It was funny i'm not a non technical person, but they have the answer like look you got bad network you got bad network you can't do much about it.
301
00:36:03.960 --> 00:36:06.240
Harvey Green: that's like not having enough memory earlier.
302
00:36:08.640 --> 00:36:13.020
Jirah Cox: What Harvey it sounds like you're saying I can't just go download more memory I don't understand yeah.
303
00:36:13.080 --> 00:36:13.470
No.
304
00:36:15.930 --> 00:36:22.680
Andy Whiteside: All right, multiple external networks is a call out here what's has that as a subset of enhanced the virtual network.
305
00:36:23.880 --> 00:36:30.990
Jirah Cox: yeah the the example given here I think it's super neat where within within a vpc right, we could define different.
306
00:36:31.410 --> 00:36:35.760
Jirah Cox: Rules to say when this virtual machine wants to go talk to an internal service like active directory.
307
00:36:36.360 --> 00:36:49.470
Jirah Cox: don't net to go do that when I do want to go out to the Internet or perhaps other targets do nat do using that to for that that traffic flow so have more control more flexibility for the way your Apps need to behave.
308
00:36:53.280 --> 00:36:57.060
Andy Whiteside: And this is all relative to building these clouds within a cloud concept.
309
00:36:58.080 --> 00:37:00.360
Jirah Cox: Right yeah fundamentally right, how do we help.
310
00:37:01.830 --> 00:37:09.810
Jirah Cox: You know it teams open up their APP store or their their hosting ability for their internal teams to go and deploy Apps.
311
00:37:10.260 --> 00:37:23.940
Jirah Cox: without needing to do sort of hard fixed allocation on real networks on real switches every single time right if you can get in software, if I could have overlapping sub nets right which is.
312
00:37:24.780 --> 00:37:32.430
Jirah Cox: very crucial for like hosting providers, you know all these problems that we have to solve that are easily more easily solved in software than they are in hardware.
313
00:37:33.090 --> 00:37:33.300
Right.
314
00:37:35.340 --> 00:37:50.280
Andy Whiteside: next section here says simplified management and automation the first subsection says vm templates for HIV Is this something that age v's catching up with the rest of the guys in terms of template management process.
315
00:37:50.400 --> 00:37:53.970
Jirah Cox: I think it even brings a couple a couple of new tricks a bit slave as well.
316
00:37:55.350 --> 00:38:06.330
Jirah Cox: The the, this is the feature i'd say for sure hop into the test drive environment kick the tires on it see what you can do, because one of the really neat tricks is the ability to.
317
00:38:07.470 --> 00:38:14.850
Jirah Cox: version your templates right, so I can like going back to like an EDC environment, I can say this is my February 2022.
318
00:38:15.330 --> 00:38:23.400
Jirah Cox: template and then I can go and deploy 100 games from that 200 games from that next one that powered on I apply my updates and security patches.
319
00:38:23.910 --> 00:38:36.720
Jirah Cox: And I check that back into the library, as the march 2022 template right so that that concept of version and templates and flagging them for availability internally for consumption, I think, is really, really slick.
320
00:38:39.660 --> 00:38:40.020
Andy Whiteside: rv.
321
00:38:41.250 --> 00:38:49.200
Andy Whiteside: And he got history tons of history with the items and templates and rolling out from templates and, yes, where does this fit in your.
322
00:38:50.460 --> 00:38:54.930
Harvey Green: Well, I mean just just makes the process easier, this is.
323
00:38:56.370 --> 00:39:01.320
Harvey Green: A new set of features kind of what Dan was talking about that we can use to help with.
324
00:39:02.580 --> 00:39:14.430
Harvey Green: Some some future os deployments that are outside of the outside of the purview of you know other imaging type software, this is something that is built in that you no longer have to.
325
00:39:15.120 --> 00:39:21.570
Harvey Green: either go create a process yourself or they'll use some other third party tool to you know help you make this happen.
326
00:39:27.060 --> 00:39:29.850
Andy Whiteside: next section talks about multi node removal.
327
00:39:31.290 --> 00:39:41.700
Jirah Cox: So to paint a word picture here, if I had say an eight node cluster and I wanted to eject three of those nodes to go send the monitor we mission to start a new cluster for a new.
328
00:39:42.090 --> 00:39:50.340
Jirah Cox: purpose right the ability to add grow clusters shrink clusters long standing mechanics value prop that you can grow and shrink as your as your needs change.
329
00:39:51.480 --> 00:40:05.340
Jirah Cox: If I wanted to inject those three nodes before now, I would object to know date of the note seven and the notes six and when I injected node eight right i've got data on there that I had to move elsewhere in the cluster so that note eight becomes empty, and I can walk away with it right.
330
00:40:06.450 --> 00:40:18.990
Jirah Cox: So you could potentially imagine that I might have moved the same bits from eight to 972 notes six and then to somewhere else in the cluster like four times, can I kind of wasted work there, so now with this enhancement a you can tell it.
331
00:40:20.820 --> 00:40:22.020
Jirah Cox: That you can.
332
00:40:24.540 --> 00:40:31.050
Jirah Cox: Remove multiple at a time, and it also doesn't therefore do cause itself any rework there to shuffle those bits over and over and over again.
333
00:40:32.370 --> 00:40:38.790
Andy Whiteside: And i'm assuming obviously that it does the math and make sure that that's safe to remove all those and me, I didn't wouldn't be in here, but I.
334
00:40:39.030 --> 00:40:49.860
Andy Whiteside: mean that this thing is very useful because more than likely when you're sizing down a cluster cluster cluster you're probably going to do more than one node if you're retiring some older stuff are moving stuff around.
335
00:40:50.250 --> 00:40:58.590
Jirah Cox: Totally right, I mean it's applicable to what's what's now a reality for our customers right that has been with us for for long enough now on their server.
336
00:40:58.950 --> 00:41:08.610
Jirah Cox: server virtualization workloads if you had a let's say in a cluster you know running your data Center maybe you bring in the latest hottest Intel R amp D cpus maybe you need six nodes now.
337
00:41:10.200 --> 00:41:20.190
Jirah Cox: we've said, for years, but now it's finally a reality for you that's like the easiest upgrade you don't have to do at all you just add those six nodes to your cluster for like three days you're running a 14 node cluster.
338
00:41:20.850 --> 00:41:27.840
Jirah Cox: And then, what you would have used to had to do was just you know one one week just go in and check the note every day right just eject eject eject.
339
00:41:28.200 --> 00:41:39.600
Jirah Cox: Every day, with your coffee you're just rejecting a different node and eventually you're down to those six new nodes now that's one click for you, instead of eight clicks to reject those eight old nodes right.
340
00:41:40.830 --> 00:41:45.390
Andy Whiteside: There it's not have that conversation I think about the two or three times I retired of saying in my career and.
341
00:41:45.840 --> 00:41:56.190
Andy Whiteside: The joys of rolling that thing out the door, which was joy, also the fact that point, sometimes it didn't fit in and out the door or in the elevator and you're going Helen heck they get it up here I don't get it down.
342
00:41:58.740 --> 00:42:03.720
Andy Whiteside: Alright, so next section talks about storage container management in prison central.
343
00:42:05.430 --> 00:42:21.960
Jirah Cox: Well, so bringing more of the what's traditionally been a prison element function up in the prison central, making it therefore under the purview of PC are back and robust control and also decentralized right, so I can now see all my containers across all my clusters in one spot.
344
00:42:24.300 --> 00:42:26.940
Andy Whiteside: I missed the opportunity to talk about the crew that acronym.
345
00:42:27.720 --> 00:42:28.110
yeah.
346
00:42:29.220 --> 00:42:31.560
Jirah Cox: that's that was not compliant it's a four letter acronym.
347
00:42:33.750 --> 00:42:35.520
Andy Whiteside: Maybe it deserves to be i'm.
348
00:42:36.570 --> 00:42:39.240
Andy Whiteside: Harvey give me your best attempt to explain why this matters.
349
00:42:41.130 --> 00:42:42.420
Harvey Green: A best.
350
00:42:44.700 --> 00:42:45.000
Andy Whiteside: topic.
351
00:42:48.210 --> 00:43:05.340
Harvey Green: So um, I think, and I guess Jerry you correct me if i'm wrong here, I think this is necessary because we've got a lot of things that are moving to prison prison centrally managed instead of.
352
00:43:06.360 --> 00:43:12.210
Harvey Green: Prison element direct on the node, and this is just helping move into that direction, a little easier.
353
00:43:12.630 --> 00:43:21.240
Jirah Cox: yeah I got a convenience enhancement right like bring bring more of the functionality into into the browser tab you're already logged into.
354
00:43:21.570 --> 00:43:21.930
yeah.
355
00:43:25.050 --> 00:43:30.180
Andy Whiteside: Alright, enhance maintenance mode how much can we enhance the maintenance.
356
00:43:31.980 --> 00:43:37.830
Jirah Cox: So another one that historically has been kind of a seelye operation now it's in the in the gooey there.
357
00:43:39.600 --> 00:43:50.190
Jirah Cox: That but also bring some new tricks to it that, when you say I need to you know change out and more memory to these nodes because we tried overcome it and we want to go back to.
358
00:43:50.910 --> 00:43:58.350
Jirah Cox: adding more memory, to the nodes then when I take that data into maintenance mode and then take it down and add dims to it and then bring it back up.
359
00:43:58.650 --> 00:44:11.280
Jirah Cox: It also remembers wireless vm lived and then moves those Games back onto the note that they belong on to or belong in air quotes there to restore data locality right, so that we don't have to chase them around within the cluster to re localize more data.
360
00:44:12.900 --> 00:44:19.080
Andy Whiteside: And just another example of intelligence in efficiencies through you know, continue to evolve, the software.
361
00:44:22.170 --> 00:44:24.930
Andy Whiteside: and explain enhancements.
362
00:44:26.250 --> 00:44:30.930
Jirah Cox: So X played not not a household name just yet really deserves to be.
363
00:44:32.190 --> 00:44:43.380
Jirah Cox: Our low code no code automation done right within prison central So what if you want to change together a bunch of actions like at you know, on Friday at 10pm take this vm down.
364
00:44:44.490 --> 00:44:53.040
Jirah Cox: Add maybe grow the hard drive or add some memory to that virtual machine power backup send me a team has message or slack message, or maybe a whatsapp if you want to.
365
00:44:54.090 --> 00:44:55.140
Jirah Cox: Call some APIs there.
366
00:44:56.580 --> 00:45:03.810
Jirah Cox: Let me know that you did it right or your email the application owner to say i've completed your change request for your virtual machine a crowd request, if you will.
367
00:45:05.190 --> 00:45:17.340
Jirah Cox: that's all existed natively within within prison central so we're making that functionality smarter here with new things that it can do so, adding here some look up actions, where we can look up this as a cluster a host or a virtual machine.
368
00:45:18.540 --> 00:45:32.280
Jirah Cox: So if you, you know that that provides all kinds of accessibility for like if you had a random IP address you can now call a playbook to say look up this IP address find the matching virtual machine and then sending back the vm name or convert it somehow.
369
00:45:33.480 --> 00:45:40.770
Jirah Cox: But then also more fun send those alerts like I said slack teams, you know if you know the whatsapp API you could probably code that yourself.
370
00:45:41.490 --> 00:45:56.880
Jirah Cox: Now natively send a pager duty right so for your OPS teams, you know that want to get alerts perhaps that match a certain level of severity right if it's if it's the the red klaxon alert that I want to get that on my phone by a pager duty that's a native integration now.
371
00:45:59.280 --> 00:46:01.260
Andy Whiteside: You guys in tell me we're going to talk about automation.
372
00:46:03.570 --> 00:46:04.500
Jirah Cox: trigger warning Randy.
373
00:46:04.830 --> 00:46:12.840
Harvey Green: I was gonna bring that up for first you allow them to overcommit memory now you give them an automated tasks to shut a machine down and just add memory.
374
00:46:13.440 --> 00:46:21.720
Jirah Cox: Well, the way I the way I describe it to admins is like if it's a vm that that that you do something on add add cpu memory, whatever add a new desk.
375
00:46:23.070 --> 00:46:32.790
Jirah Cox: And you only do it during a maintenance window and it's only Fridays at 10 well, you could you could simply be awake Friday at 10 on the laptop on the vpn and make the change.
376
00:46:33.240 --> 00:46:38.340
Jirah Cox: Or, I would rather you know do that Thursday morning, you know nine in the morning program all the changes.
377
00:46:38.760 --> 00:46:51.000
Jirah Cox: use a repeatable run book for that, instead of trigger based on time to say hey at 10pm when you automated system are awake you go do this for me and just email me if there's any problems with it, otherwise you know don't bug me.
378
00:46:54.180 --> 00:46:58.500
Andy Whiteside: yeah you guys like if I was still running the world you'd have to like crank the thing on the front of Carter started.
379
00:47:02.910 --> 00:47:05.790
Andy Whiteside: Imagine how much bonding between husband and wife that created.
380
00:47:08.040 --> 00:47:09.870
Harvey Green: Oh, my gosh i'd rather not.
381
00:47:11.790 --> 00:47:14.700
Andy Whiteside: No i'm just kidding i've gotten used to punch in the button.
382
00:47:16.710 --> 00:47:23.370
Andy Whiteside: Well, like always it's pretty simple to go kick the tires with this thing called test drive I assume it's reading.
383
00:47:23.790 --> 00:47:33.840
Andy Whiteside: Well they're just says right there and test drive today so everything we've talked about here can be played with in a matter of moments just by going out to the new tactics test drive site.
384
00:47:34.530 --> 00:47:39.300
Jirah Cox: yeah I think it's I think it's super exciting it's you know mix of major and and.
385
00:47:40.590 --> 00:47:49.110
Jirah Cox: You know more visible and less visible enhancements right kind of a mixed bag they're very well rounded but you know throughout it all the key is.
386
00:47:49.590 --> 00:48:00.570
Jirah Cox: Just software that makes your structure get better over time right so it'd be like your engine X cluster does more for you tomorrow than it did yesterday and runs better runs faster over time, I think that's the coolest thing over.
387
00:48:01.620 --> 00:48:07.620
Andy Whiteside: Well, and for people that have been around me lately i'm constantly talking about digital transformation what phase you're in because it's never ending.
388
00:48:07.830 --> 00:48:13.350
Andy Whiteside: And because integrity integrity we're always talking about digital workspace fronting that digital transformation.
389
00:48:13.620 --> 00:48:24.360
Andy Whiteside: To I was doing that to you and Harvey jumped on a while ago I showing you our workspace right you've got this amazing digital transformation thing have any new tactics, I would argue that your digital workspace that front ends it isn't.
390
00:48:25.410 --> 00:48:34.680
Andy Whiteside: I mean, maybe you have another one one I didn't see but the one that I saw was very lacking in terms of all the things that that can be leveraged for to go along with that digital transformation.
391
00:48:35.790 --> 00:48:37.380
Jirah Cox: We don't we don't have different was banks don't we.
392
00:48:38.430 --> 00:48:38.700
Jirah Cox: yeah.
393
00:48:39.960 --> 00:48:42.900
Harvey Green: So so Jerry you did all this in a weekend you.
394
00:48:43.530 --> 00:48:45.960
Harvey Green: have yourself a raise that looks like a fancy charity oh.
395
00:48:45.960 --> 00:48:47.850
Jirah Cox: Thanks, but we had a long weekend.
396
00:48:52.710 --> 00:49:00.930
Andy Whiteside: Alright guys well we're out of time i've got another podcast to do, but I appreciate you guys joining and we'll get this posted in the next 48 hours.
397
00:49:02.010 --> 00:49:02.790
Harvey Green: sounds good.
398
00:49:02.910 --> 00:49:03.780
Jirah Cox: cool talk to.
399
00:49:04.470 --> 00:49:04.650
me.