1:03:26 Webinar
Ask Us Everything About Storage for Databases
This month we’re diving into databases—and how Everpure (formerly Pure Storage) helps infrastructure teams support modern, mission-critical workloads with less risk and operational overhead.
This webinar first aired on 19 March 2026
The first 5 minute(s) of our recorded Webinars are open; however, if you are enjoying them, we’ll ask for a little information to finish watching.
Click to View Transcript
All right, well, good morning, good afternoon to everybody. Thanks for joining us on yet another Ask Us Everything. I'm Don Foreman, as you know, technical evangelist here at Everpure, and today we've got a really good topic that we've teed up for you guys. We're gonna be talking about all things databases, right?
All things structured data, which is kind of one of our big shining pieces here at Pure when you think about all the things that we have done with our technology. And something came up the other day that I was watching that made me realize, wow, we really need to start thinking about structured data again. It was Jensen Huang's keynote at NVIDIA GTC, where he even mentioned
that while s- unstructured data feeds and feeds the models and all of that there's a massive amount of classification that happens behind the scenes. And what is classification? Well, that's turning things into structured data. So I think the conversation we're gonna have today is gonna be very relevant to not just
all of you people out there who run your Oracles, your SAPs, your, your SQLs, all of those things, but forward-looking stuff as well. And the best part is, is I'm joined by two of our in-house experts, Ryan Arsenault and Anthony E. Nocentino.
You guys, y- you know, obviously we've known each other for a while, so this is gonna be a really good session. I'm gonna step out of the way. Anthony, you wanna start, introduce yourself, and then pass it over to Ryan, and we'll get things going? Sure.
I do wanna pull on that thread, though, for a second, that you were talking with, structured data, and I think what I'd love to talk to people about is that's the representation of your business right now at this point in time, right? And that's one of the key things. I think that's one of the things that we see AI h- hopping onto in that,
in that keynote there. But hey, everybody. I'm Anthony Nocentino. I'm a senior principal field solution architect at Pure. I specialize in, in database systems, most specifically SQL Server, but pretty much anything that stores data, in a structured way.
And I'll take it over to Ryan. Say hey. Yeah. Hey, everybody. Ryan Arsenault, principal field solution architect here at Pure, Anthony's work wife, as people like to call us. I almost spit my coffee out on the screen. That would've been, that would've been epic.
I've been with Pure seven years. I cover I'm also part of the database team here, but I focus really on SAP. Awesome, and guys, I, I think it's gonna be great today to just give a lay of the land of y- you know, the state of structured data, the, the state of what we've been doing at Pure, obviously, to continually improve the entire structured data experience, and give some
hints on where our brains are as we move forward. Especially to the point that we just talked about, that AI is going to bring some structure, more structured data into the mix, and Jensen's keynote, I think, So I'm gonna turn it over to you guys, and I'll just fold in Q&A as we go along. Cool, cool. Yeah, I guess let's pull that, pull that
thread again a little bit, the AI angle, specifically in structured data, and I think we see what the, the major database vendors are doing in market, or more enterprise database vendors are doing in market, Oracle, SQL Server, and, and SAP, right? The folks that are bringing vector data stores into these, these kind of enterprise
workhorses that have powered these businesses for the last 40, 50 years, right? What do you think, Ryan? Yeah, I mean, it's super smart, right? Like you said earlier, that's where the business data lives. Instead of moving the data to the vector database, why not move the vector capab to where the data is, right?
Into these enterprise systems like SQL Server and Oracle. So I, I think it's super smart. I love the fact that everybody- Yeah, it was I'm a- Sorry to interrupt. I love the fact that everybody's slapped AI on the end of their database engines now. Like, Oracle did it. Yeah.
I think SQL did it. Like, you know, like I can put this- Right on my cup, I could put, "This cup powered by AI," you know? I think But your point does make sense. Don Foremen. Don Foreman, powered by AI. Right. Then your stock price goes up once you do that, right?
Right. Right. Yeah, I think one of the, the I stopped When I start talking about vector to customers and to our internal folks to show them the bits and the bytes. Said, "This is a vector. This is what a vector index is. This is what an embedding is," and recently I stopped doing that.
I started showing the outcome. I showed a web application where you can natural language interact with structured data, right? So for years as developers and, or even as consumers in an application, you know, if I wanted to buy a bicycle, I'd check a box, bicycle.
I'd check a box, yellow. I'd check a box, 24-inch rims. And I'd have to build this structured thing that as a developer I'd have to turn into some structured query to go extract that information from the database. Now, with AI, or, or more specifically, semantic or similarity search, I can just
write in a natural language way, "I want a yellow bike with 24-inch rims," hit Enter, and the database is gonna go figure that out without a complex coded structure. Yeah, I think that's, that's really where SAP's heading, too. You know, SAP, as you know, Anthony, is more than just a database. It's a whole- Is it now? It's a whole suite of applications.
We're gonna riff, people, so just let us do it. Um- Your suite. It, it, it SAP is really kind of replacing their UIs with what they're calling their Joule AI agent, right? So no longer if I'm an AP person working business and I need to go pay invoices in the morning, no, no more am I going to a GUI and saying, you know, searching for all the
invoices I want. It's just show me all the invoices I need to approve today. And then, you know, if you think you can approve them, approve these invoices. Boom, right into a, right into a chat ag- AI agent and away you go.That's a hell of an observation because that actually came up last year at Pure//Accelerate in the navigators
discussion, where we were talking about AI Copilot in Pure1 and how it's able, through natural language, to get you to do whatever you wanna do as far as sleep management is concerned, or anything through Pure1. And the question did go out, is what is the relevancy of a GUI over time if an AI chatbot becomes your interface, right?
Now, granted, it's not gonna be wholesale replaced, and it's, it's always going to be there, but we are advancing how people interface with data in general when we're talking about this chatbot thing too, right? Yeah. I've been s- I don't wanna say cheating, right? But I've been experimenting, with database performance testing using Copilot, where I
have my structured tests, but what I do is I take the output of tests, put them into Copilot, and it tracks the iterations over time- as I make changes to the system that I want the performance test. So I'll go and I'll mash a button or change a configuration setting in my system, I'll run the test, it'll grab the output, it'll build a nice comparison of each run for me so I don't have to go and copy into a spreadsheet and
make a chart and do all the things that I'd have to do before. So even in something that's very practical for me day-to-day, I'm leveraging that to just kind of iterate quickly and analyze, a S-A set of data quickly to make decisions, and so that's a practical way that I'm using that every day. Copilot specifically.
That's super smart. I know, thank you. Thank you. You should share that with the team. Yeah, well, it's I, I mean, you know, the whole advantage of AI, right, is pattern matching at massive scale. And- Right I think that's the advantage you gain from this, is that you offload the
pattern matching to something that's much better armed for it than your brain, right? Not that you aren't a smart guy, but it sees in dimensions that you just, y- y- you l- you reach a limit in your brain, obviously. So th- there are a lot of advantages, I think, to all of this coming out. But, like, I, I don't wanna get us off track.
Obviously, you guys have a whole litany of things you wanna cover, so, you know, I say let's dive in and, and fire away at it. Sure thing. Yeah, and if anybody in the audience has a, has a database question they wanted to get us, get into, pop it in the Q&A, and we'll definitely bubble that up.
But I do have a, a seed question for Ryan because- Ooh to kind of start the conversation. Ready? Uh-oh. So Ryan, you get to work with the biggest SAP customers on the planet, right? Running on our biggest storage device, the FlashArray//X. And so when you talk to them about consolidating workloads, like, how do you kind
of establish or explain the difference between th- the performance and scale that they would expect on FlashArray and compare that to, let's just say, legacy storage? Legacy. Legacy all the way, right? A- and before you do- Mm-hmm actually, Ryan, before you do, we actually have a question that came in related to what you're about to say.
When you're running big business critical databases, what does Everpure actually change for the infrastructure team day-to-day? So- Beautiful man, this could be perfect. Beautiful. You guys are insane. Love it. I, I really, I, I personally really focus on how latency behaves as workloads are stacked on the array,
rather than just focusing on, "Here's some hero numbers that we've generated in a perfect environment," right? So a lot of my customers from an SAP standpoint, you know, a lot of them are just, are not mixing workloads, right? It's, "This is my SAP, I'm buying for this business unit, and all the SAP's going here."
But some customers definitely do mix their workloads, you know, think mission-critical databases, VDI workloads, file-based and, and stacking these onto XLs and, and continuing to get sub-millisecond performance, even, you know, 150, 250 microsecond range. You know, for me from an SAP standpoint, I, I think of customers who do maybe a lot of M&As, right, where you're, where you're buying,
you're purchasing a company who has their SAP platform, has their own data center, their own storage, and it's, and it's really hard to consolidate at the business application layer 'cause they may do things differently and, you know, the business application does what it does in a business process. So they s- look at consolidating at the storage layer, right, to at least get those
efficiencies in the data center while, they can work on what that business transformation can look like, and that FlashArray//XL can handle, handle that scale. And Anthony, I know you guys did some testing from a, from a SQL Server standpoint. I think it was HammerDB using Tproxy, right, where you guys did- Yeah, Tproxy its new, its new order creation simulation, which was What was the number?
2.2 million new orders per minute? We hit 1.2 million IOPS in five rack units on an XLR-5 at just over 2.2 million, I think it was transactions per minute for that one. That's correct. And yeah, I think the key, the absolute key is the simplicity. Like, we didn't have to go and put on, like,
our lab coats to make the FlashArray that fast. We plugged it in, it went fast, right? Yeah. Yeah. And we actually had a customer repeat that same result. Like, first we, we slid the paper across the table and said, "It's this fast." And they're like, "Sure, we wanna see it with our own
eyes." and so they plugged the array in. Within five days, you know, they had some other infrastructure shit they had to sort. And then when they finally settled down to get the load test up and running, they went from plugged in, five days, 1.2 million IOPS, and, and, and no tuning on the array side. And, and that's, when you talk about enterprise w- managing tens and hundreds of
arrays, that simplicity provides so much value because you can shift your focus on whatever the next problem is in your data center, not making sure that your storage is tuned properly or what knob or button has to be mashed to keep it happy and healthy. Right. Right. And I'd be remiss if I didn't mention, you know, our workload planner in, in Pure1, right?
As you're consolidating- Yeah very, to say, "Hey, if I, if I take this off of here and put it on this XL, what's it gonna do to the XL?"Right? It just makes it- Right very, very, very simple. Yeah. So I think- Yeah, so- You know, a, a couple of things you're saying here to answer that question that came in, there's, it's, it's a
mishmash, right? It's the simplicity of plugging in and getting a million plus IOPS without putting nerd knobs through the motions. Right. There's visibility, obviously, that, that needs to happen, projecting what workloads might look like if they were to be implemented in. And then I, I think the biggest one is something you said earlier, Ryan, is we're
just gonna keep stacking stuff on the XL, assuming that it will just keep with it, right? And this plays into something in a blog that's going out today that I wrote about, that is dispelling what enterprise ready means, right? That we hear enterprise ready all the time, and it's mostly thrown out by legacy providers who say they just don't have the proven record or, or some made up thing, right?
And no, enterprise ready is you keep stacking stuff on your array of mixed bag things, from databases to RAG queries to anything, and it just keeps up with you behind the scenes, right? I think that's a heck of an observation that ultimately is the key to answering that question, is y- you don't have to think very much about it.
I A- and we're not trying to insult you that these knobs, you know, didn't matter before, but we want you to think about the bigger, important things, right? Totally. Right. Like, focusing on the actual data. And I think that's the other dimension in, in large scale business critical databases. We've talked to customers that have 100, 200 terabyte databases that they quite literally
just can't backup, right? Because they got to pick up, sc- pick up 100, 200 terabytes, write it somewhere else, that's cool, then I gotta restore it. That's hard, right? And so then you have- we can provide ways to help customers really solve those problems.
So FlashArray with snapshots, I'm a huge fan of snapshots. If you have a 100 TiB database, even with the, the rocket fast things that we have on the truck at, at Pure, at Everpure, for recovery, instant is a recovery pattern on FlashArray with the snapshot. And so be- imagine being able to look your business in the eye and say, "I can return
this business back to full function instantly in the event of a failure a powerful thing to have in your, in your tool belt. In addition to that, when you look at kind of the FlashBlade side of the house, our solutions team, Mark Wilkinson and Andy Brown just released a white paper where they're able to backup at 1.8 TiB a minute and restore at a TiB a minute.
Dramatic pause intentionally, right? Yeah. So in addition- Assuming your network can, assuming your network can keep up with you, right? Oh, yeah. It literally just becomes a network engineering exercise, right? The, 'cause the box is so fast.
But the idea here is when we're building enterprise strategies for data protection, to really hit those aggressive RPOs and RTOs, the answer is both, right? Having on array snapshots for instantaneous recovery, and also having the ability to write RMAN or native SQL Server dumps, or what's no, I'm not even joking this time. What's, what's SAP use for backup, Ryan?
They, they, they have a, they have an SAP HANA backup utility as well, yes. Okay. It's a big boy- So it's- big boy database. It's a big boy database. Yeah, so being able to backup and restore in a FlashBlade, or more s- more along the lines of information lifecycle management. Like, I think when we're talking about recovery, there's different recovery patterns
that you need to support in an enterprise. One, keep the business online. Two, information lifecycle management, regulatory compliance, maybe troubleshooting. Maybe I'm a developer and I found a data anomaly, and I need a restore from seven weeks ago before I did my code push, right?
Those kind of things was the answer, to enterprise recovery strategy of both snapshots and clones for instant, but also native dumps, RMAN, big boy backup utilities at SAP, dumping data into a traditional backup scheme. And we've talked about this before, right? What is, what is a DBA's biggest fear, this overnight backup job is gonna be
running when they wake up in the morning and come in, right? And- Right and what else happens in these systems overnight, right? It's not just a backup job. There's probably batch jobs going to prepare for morning, morning reporting for the business, and those things can't affect each other.
So DBA wants to get that backup done and off the storage system as quickly as possible, and we enable that with Pure Storage. And, and, and restoring, God forbid to ever restore that database, right? FlashBlade, as you just said with the, statistics, just it, it gets it back quickly as humanly possible, to get back up and running.
Yeah, one of the other outside- Yeah angles to talk about, too, while we're talking about snapshots, is SafeMode as well, right? Like, the DBAs are not immune from ransomware attacks, right? So the snapshots of unstructured data are just as important as they are with structured data as well. And, you know, you bring SafeMode into all of
this, the database recovery in the instance of ransomware even becomes as well. So it's a multidimensional value that we bring to the table at Everpure related to getting your stuff back if you need to you're talking 100, 200 TiB databases. Don, I support a 53- Yeah year-old piece of software that is still patching security holes
every week. And it's also usually the lifeblood of a, of an organization's money. You think- Big if a company gets ransomware on their SAP they're not gonna pay? They're absolutely gonna pay. So this becomes critical for them to know that if and when, knock on wood, right, a, a
ransomware attack happens, they can recover and they can recover quickly to keep the business moving. Wait, 50, 53 years old, like, h- was it written on punch cards? Yeah, yeah. It was. It was. Yeah? It sounds like- Well, how do you ransom, how do you ransomware a punch card?
Do you, like, spill a beer on it? Like, what's going on? Yeah, r- yeah, right? Exactly. I, I was gonna suggest that we're in, like, database four territory here, but- that might be newer than, what, 53 years. We're going old school here, folks.
Yeah, that's really old school. Oh, man. So going back to the original talk track of everything, you know, what's the best way to start, you know, trickling down recent things that we've done related to database performance and longevity and health and all of that. Obviously, we've got a hardware thing out
there now with the FlashArray//XL 190, but Yep other things we've done inside of Purity that I'm sure are worth talking about. Ooh, okay. My biggest favoritest thing that's come out in the last year, and it's gonna sound not super sexy, but it is, tags, right? So we've added tags to nearly every object inside of Purity.
So volumes can have tags, protection groups can have tags, discrete snapshots can have tags. So when I take a snapshot, I can add tags to the snapshot, which also means I can tell the system what's in the snapshot. So no longer is it, you know, protection group LMNOP.247 or whatever the protection
snapshot number is. I can actually add some intelligence to it and query the snapshots in my environment. Cause as a DBA, I'm not looking for protection group LMNOP217. I'm looking for this database from this server, right? Yep. And so we can add that information into the
snapshots and go and search the fleet for that information. So last summer, your partner in crime, Mike Nelson, sent me a DM- and, he's, he- all the DFM was, was the release notes to REST API at 2.4.4 or 244, and the words, You're welcome." And I read, I knew ex- I knew exactly what I was getting. I was, "Yes." And then I wrote a bunch of code that did exactly that, right?
244, adding tags to the snapshots, and so now I could go into an array and search, I want that snapshot 'cause I could, could just go ask the array for the database. Now, about two weeks ago, Mike sends me another email with some release notes, "You're welcome." And it was REST API 251, which for a Fusion enabled fleet allowed me to ask one endpoint to search the whole fleet for a snapshot.
Ooh, ooh. So now Right? Right. Ooh. So now I have a snapshot catalog, and so I can have the intelligence of where my data is in my environment without having to understand a volume or a protection group or even where it lives in an array. I just ask the fleet, "Where is this database snapshot?" Super powerful, favorite feature.
Yeah, that's, that's- No, as, as these snapshots move around, the tags move with them too, right? 100%. Yeah, 'cause they're object-oriented, right? Yeah, and that's, a- and, and I think that's pretty powerful because that plays into STaaS common theme that, that I've heard is when something goes wonky, when something slows
down on the Array, I, I don't know what it's tied to. I, I know the volume that it's on, I know loosely- Right where it is, but I, I, there's no chain of custody basically, right? So this tagging thing, man, opens up a whole litany of possibilities, especially when you can query it across the fleet now.
Yeah. Yeah, and in, in the Fusion context of things, one thing I'm excited about is the, we're, we're gonna lean into, we're kind of control freaks. I don't know, the storage folks- No have probably experienced this with us. No, no, no, no. I don't believe that. But, with Fusion with presets, right?
So now I can define what good looks like in a fleet for a database server type of database server, Oracle, SQL Server, SAP HANA. And then I can also define its protection strategies, its performance boundaries, you know, gold, silver, bronze, intersite replication, all those goodies come with that, and I've, I've been really enjoyed being to deploy workloads with presets, with that.
So what are your thoughts on that, Ryan? Yeah, I mean, thinking back to my M&A conversation earlier, right? You, you take over this new SAP HANA environment and it's completely different have in your production environment. So presets are a great way to standardize, you know, across my mission critical, maybe
business-critical, maybe business-supporting environments, as well as your, you know, your non-prod, your dev, your test environments. I have some customers where their dev environment is more important than their test environment, so they have a different SLA on that because a lot of code is stored in dev and then it moves to test.
The, the dev is the, is the source code for your future releases. So put a, put a snapshot policy on that, and presets just make it so much easier to just standardize across the fleet, across your platform, and protect across, across all those different workloads. Yeah, 100%. The other one, the, the presets that is
interesting, especially to DBAs and storage pros, 'cause we've all been burnt by this. How many of us have gone into, a system and added a data file, put it on a new volume, and forgot to put the new volume in the protection group? Right? And then when I come back and take the snapshot of the protection group, I am very, and I need that for a
restore, I am very sad, right? So with Fusion presets, if I depl- if I create a preset and I describe a disk topology in there with a protection policy on that, and I deploy, if I, on the deployed workload, if I come and I add another volume to it, it's automatically gonna inherit the performance policy and the protection group policy, making me happy over time.
Because we've all done it before, and this kind of really reduces a significant amount of risk to the business and to us, honestly as, as individuals, 'cause we don't want this to happen to us, right? And so it's a fantastic kind of, I don't wanna say hidden gem, but definitely a, a very discreet problem that we've all had probably in production systems over the years.
The heart-dropping moment when you realize that you don't have a snapshot- Yeah that you need to restore, right? Yep. Exactly that. Nothing worse than that feeling in your stomach when data loss is possible. Mm-hmm. Yeah, we've all done it.
I remember the very first time I, I dropped all the rows from a quarter of a million row table and, I kind of just backed away keyboard, waited for that funny feeling away in my stomach, right? Mm-mm. Ca- then I came up with a plan and, like, got back to work. But it's, yeah, we've all done it, so.You didn't go to your Google
Drive and find your resume? Yeah, that's ri- that is a resume-generating event, isn't it? That could be. For sure. For sure. That could be. 100%. So outside of Fusion, what, what else have we been doing inside of Purity hardware, everything along those lines to improve the
database experience as well? Cause I know database people love their performance numbers, so we obviously have to brush up against that, right? To flex and, and show the world we know what we're doing, right? Yeah. Do you wanna grab this one, Ryan, or me go?
No, go ahead. Cool. I am, I'm a performance junkie. I like to make things go fast. I like to make the green light in the data center blink as fast as possible. That's, like, one of my It's like scratching an itch.
And we, one of the things I spend a lot of time in with is, our open metrics exporter, right? Which used to run as a external container, where you could go and, and wire that up, and hit the REST API, and expose the metrics to whatever system you want. Usually, I use Prometheus and Grafana, to build performance dashboards for what's going
on inside the array. It allows you to get a real finite on performance metrics, where when you look at things like in Pure1, there's some roll-up that occurs, so maybe you need a more or even longer term, representation of the performance in your environment. So we took that, we put it inside the array, so now there's a Prometheus
inside the array today that you can go and interact with. And I love using that to build correlated dashboards with the database performance metrics. 'Cause we've all had the DBAs come to the storage folks and be like, "Your storage is 100 milliseconds," and the DB- and the storage folks come back and it's like, "No, it's not, it's .1 millisecond," and then there's
something in the middle that's causing queuing, right? And so building these correlated dashboards can help you get to an answer faster what's going on inside of it, inside your applications and inside your storage. So love, love using that tool. The, speaking of snaps, we just had a question come in that Ryan just answered,
gonna throw it out there- Pure in case anybody's curious, is, "If I do take, take a snap on one array, can I move it over to another array and r- restore it?" And yes, you can, obviously. Yep. I mean, the assumption there is there, there has to be the bones of what you're restoring, but y- you wanna talk a little bit about that, Ryan? Yeah, and it's block-to-block, right?
You, you can't go block-to-file. But if, y- you know, if you have I think a great example of this is a refresh, right? A lot of customers are running their production on one array, and maybe their dev test on another data center on another array, and you wanna be able to move business data over to your test environment or your dev
environment to be able to recreate or fix certain business processes from production environment. Yeah, you take a snapshot of your production environment. Our data's mobile, right? That's the number one thing I love about it, is how mobile our data is. And that, I think that's probably gonna lean into some cyber resilience
questions here in a minute. But the way, the ability to take your data, your snapshot of your data, and move it somewhere else, and then restore a clone as many times as you want, it's, it's pretty life-changing. And it's what I talk, I know you do as well, Anthony, what I talk to DBAs around the most, is you don't need to backup and restore to be able to move that data.
You can do it in milliseconds- Right a snap and, a snap and, and, and And, and a lot of times in that process, you don't even need to create your database, right? You can have your database up and running and, and take a, a crash consistent snap that will start up on the, on the other array. Yeah, one of the things- The, the way that I- Go ahead.
Finish your thought, Anthony, and I'll throw mine in. Sure. Yeah. The way that I, the real special sauce, in addition to the fact that a snap can be available somewhere else, is this. If I'm replicating data between two arrays, that, that first replication is gonna be the data reduced set between the two.
So let's say, for easy math, four terabytes, right? And I get four-to-one data reduction. We're gonna move a TiB to the, a, a real network transfer stuff between two arrays. But then when I come back and take the second snapshot, what I'm gonna do is just send the change blocks to the target, but I'll have two completely unique snapshots
on the target, right? So those change blocks has just changed from the last snapshot. So I see something about the snapshot frequency. We can certainly have it on a schedule as low as five minutes, so you're gonna make little rocks out of big rocks, even in a, a highly changed velocity database.
So I'd be able to have those across two different arrays, and I can bring this database online instantly. There's not four terabytes of weight over here. It's gonna be snap clone, consume data just in a couple milliseconds. And when we look at things, like, one of the other things I've talked a lot, that's, that's
a fantastic DR scheme, but you could also push it up into your applications. You could seed SQL Server availability group replicas with this technique. You can seed Data Guard replicas with this technique. Yeah. Right? 'Cause one of the harder things that we have to do as data professionals is, so if I have a DR site and I have a 50
database at DR site and I fail over, when I fail back, I've gotta bring all 100 TiB back to the original site, restore that, three, four, five days, however long that takes, right? Spend my time watching backups restore for however long, and then bring all my application level replication backup. On Pure, we can do all of that with snapshots.
Reseed replicas, bring everything back online. So imagine that fail over, fail back. That fail back, again, change blocks come down, snap clone, reseed replicas go. And that becomes a seconds operation instead of days. And huge value because you can put your business back in its regular HADR posture
almost instantly, right? Yeah. One of the- You bring up a good, a good point. Sorry, Don. You can mix and match, right? Between software replication and storage replication- Yeah to get- Yep the, the best, the best SLA for, for your business.
I, I helped a- Yeah very large customer j- do just that, right? They had an ActiveCluster pair for their large, very large, SAP database, using then SAP HANA replication to send it to another data center where they had another ActiveCluster pair. They also snap and move those snaps over for when, you know, high transaction times,
uh-Let's be honest, SAP's replication sometimes starts to slow down a bit. If it slows down too much, there's no reseeding, just like Anthony said, reseeding that replica with that snapshot so you can catch it back up- Yeah so it can then continue on. It's, it's, it's really, really cool technology.
Well, and you, you opened up a Pandora's box that you predicted, Ryan, with, cyber questions coming in with data mobility. So we did get one in that was like, where do you think the line lies between effective AI access to data and the risk of losing control over it when the infrastructure becomes too complex and all of this mobility is around there?
You know, what where are the gaps and the holes? I stumped you. Dramatic pause, Ryan. Yeah, dramatic pause. I'm thinking about that one. That's a very dramatic pause. Go ahead if you got it, Anthony. Well, I tell you what, while you p- while you ponder that, I will throw one extra thought in
about the snapshot thing- Sure while Ryan chews on it. Um- Mm-hmm the other example of snaps leveraged effectively is something that out when we talked about realms in FlashArray as well, that your primary database can exist one place in a realm, and a completely different realm could be your test dev that uses a snapshot of it, right?
And the system will do the QOS thing and keep things under control with the two being kind of related to each other. So I think there's a ton of advantages as well with working with snapshots in that kind of sense. Yeah. You ready for this one, Ryan, or you want me to grab it? 'Cause I got some thoughts.
Well, I, I think the question may be heading towards, you know, how do we, when, when your data is no longer secure, right? And what kind of- Right predictive, analytics does, does Pure do- Mm-hmm to, to, to look at your data? And I think the number one thing that, that comes to mind is when your, when your database,
let's say your giant 10 TiB database is very happy getting 4:1 data reduction, and then all of a sudden it's down to no data reduction, right, 1:1 data reduction, let's say. One of two things happen. You've got someone in there, a bad actor, who has taken over your data, right, and, and is now, about to give you a ransomware attack.
Or you've got a DBA that didn't know any better and just turned on, some sort of security measure and, and, is now, able, is now, now not knowing that that's gonna now take up a lot more storage on, on your database, right? So I think Pure is, is constantly monitoring that. We're gonna let you know. If your, if your data reduction suddenly goes
down to 1:1, somebody did something to your data and, and you may wanna go check it out. Yeah, that'd definitely be a concern, someone turning on encryption, unknowingly. We've all I think we've been there, uh- Yeah having those conversations on where encryption should live in an environment. Yeah. Looking at this, through another lens, I think if we're looking at kind of
and access to data, I don't think that that's problems unique to AI. I think that that's really just data governance in general as you move data between systems, right? Whether it be analytic systems, non-prod systems, but now we also have this other thing that we're feeding with our data, which would be any sort of AI environment, rack, large
language model, whatever. And so it- it's the same problem just with the newer, sexier marketing campaign around it and, and how we manage, an environment and its data. Another lens to look at this through is one of the reasons we, like, I specifi differentiate between enterprise AI and kind of the data scientist level of AI, is if you look at Oracle and SQL Server and SAP, they've been doing with AI, all of the
all built into the core database engine. So all the enterprise security that you have surrounding that database today protect the AI elements as you expose that from your application. So if you are using transparent data encryption, or if you are using row-level security, or Active Directory, or Entra ID, Kerberos, or whatever it is to secure access
to your data, that's where all that tier 1 data is as you bring these AI capabilities into your environment. Now, when you do things like generate embeddings for the data that are inside of these enterprise databases, then you also have to think about, oh, where's that data going when I stick it in a GPU to embedding and stick back into my KVDB?
So I might wanna actually have physical control of those GPUs in my data center versus having that go out to the internet to generate embeddings against an open AI consume the embeddings back in my environment. And so you have to think about what these look like within your enterprise architecture, and also how that bumps up against your internal compliance, regulatory
compliance, situation, right? Could that be considered something at the networking level too? I was doing some reading on Nutanix's AHV and how- Mm-hmm they're using Flow now, which is their micro-segmentation to manage connections into Bluefield, right? Which is the- Okay DPU stuff, right?
Mm-hmm. So- Right part of this is obviously your data to prevent a violation, but you also kind of have to have the network under control too other than otherwise, it's a free-for-all. It's just another door that it can go out basically, right? Yeah, 100%. I think data, ever since, like, the beginning
of my data career, data exfiltration has been my nightmare scenario, right? As, like, a slow leak of data just going somewhere else that I can't detect for some reason. And, and, and being able to leverage the network to put boundaries around that is, is incredibly important with what you just described.
So, yeah. You mentioned LDAP, and we did get a question that said, "When configuring for an array, what is the added benefit of doing a check peer against the CA authority versus just the LDAP? And when would you not use it obviously?" Since we are talking security here, this seems
pretty appropriate. Yeah. I think if the question, is why do we certificates against the certificate peer is because you wanna fight, you wanna trouble, you wanna prevent-Uh, the eventuality of a man in a middle attack, right? Yep. The idea is I need to be able to trust the remote certificate, that it's signed by an
authority that I trust too. Mm-hmm. Right? Yeah, and I get, I get we've all had a lot of heartburn around certificates and having self-signed certificates or internal CAs- Oh, yeah but there- there's a reason why that's hard, that's hard, because we don't want anybody to fake it.
And so we're just asserting that level of control that you'd expect, in a piece of enterprise software. And we always love the annual or the, the thing that happens every year when we forgot to renew our route, right? That's always a fun one.
That's never brought an application down that I've supported, not once ever. Never at all. Never. I, I remember route certificates were so important in VMware with a lot of things, and every- Right year people were like, "Oh my God, I forgot to renew it." So.
Every year, on the dot. Every year. Ryan, there's a question for you that just came in that looks pretty interesting. Ooh, this is a hot one. It says, "You, you often highlight snapshots as a sanity saver for basis admins.
Given the-" Yeah " transition to SAP RISE and ever-increasing automation, how do you see the role of a basis engineer evolving? So this is very conceptual. Are they gonna, you know, rather than be data cons- custodians, are they gonna be workflow architects? I mean, that is a possibility here because we are talking about evolving the career, right?
Evolving the contribution. So what are your thoughts on that? That's a fantastic question. I see it's anonymous. Was that you, Anthony? No, it was not me. Okay, good. Um- I couldn't write something that smart about SAP.
I'd just call you, right? That's true. That's true. That happens. So that's a, that's a great, unbelievably great question. So just so folks know, SAP RISE, that's SAP's managed service of, of an SAP HANA they're- they're trying to get their customers
to shift to. The first thing I tell customers when they're looking to go to SAP RISE is r- read the RACI. And it's hard. It's 1,037 lines last time I've, I've noticed, last time I looked at it, and it's, you know, what is SAP responsible for? What is the customer still responsible for?
And you'll notice, there's still a good chunk of basis level things, like really, hard things that the customer is still responsible for. Now, it becomes harder because you no longer have the access that you're used to. You don't have route access. You don't have OS access. You don't have database access.
But there's still a lot of things that you have to do at the application layer. One, one great example, we talked about test dev refreshes earlier. To request a test dev ref- refresh in a, in an SAP RISE environment, SAP will press the button which will restore your database, right? All that does is restore a copy of your production database over to your dev, over to
your, over to your test environment, and then you have to do everything after that to make it actually look like test. Repoint all your, all your connections, run the jobs that change the data from the production environment to the test There's still a lot in there that you have to do.
But what I don't want you to do is, as engineer, is become, a ticket watcher, right? I don't want you to s- log a ticket saying, "I need this code moved," or, "I need this refresh done," or, "I need this test dev, restore happening," and then you're staring at a black box waiting for SAP HANA to respond. That's something that y- every organization needs to avoid when we, when you go into a
RISE contract. I would honestly lean in heavy to, to AI right now. It's SAP's number one, number one thing that they're looking to do. There's, there are Even though they're cloud-based, there are still gonna be basis-type activities in things like their, their business development cloud, Joule Agents.
They're, they're looking for customers to be able to build those as well. I would lean in on that. And then yeah, you're gonna be more of a data person than I'd say a You know, I'd always joke that basis people are the, the, the, the most, non-predictable people ever. They're fighting fires constantly.
That's gonna start to change. Those fires are now SAP's to worry about. Start, start working on how you can advance the career at, with things like AI, with things like Joule Agents, and with things like, you know, their business develop- their business data cloud, which is the future of their, their AI platform. So it's I, I see it as an opportunity, to be honest, but don't, don't
become a ticket watcher. That, that's- Is it really- the number one thing. Is it that RISE is more of a, a STaaS platform than it is a PAS platform to really- It's a STaaS, it's a STaaS Lite, right? Yeah. We're gonna do the SAP's gonna do the bare minimum to keep your SAP, your SAP environment up and running. Um- Okay and, and you still have stuff that
you have to do, an everyday task to make sure that system is performing the way perform, is available- Interesting the way it needs to be available. Um- Okay it's, it's an interesting You and I, we'll talk over a beer about this one at some point. Well- one of the things, one of the things you brought up, Ryan, was something that I had learned years ago when I was an SE at a
company you and I worked at, and I was focusing on service providers. And the idea of moving to a service or a cloud provider is giving up a level of span of control, right? Right, right. That you're making certain assumptions that the provider will put the same amount of emphasis and importance on the things that are
important to you, right? Mm-hmm. But that's not really possible because they're broad-brushing it across a much bigger group of customers, right? So that, that fine print does matter, obviously. And again, I, I think to your point is, y- your house needs to be in order before you
just heavy lift it somewhere else, right? Right. Cause otherwise you fall prey to something we used to say in that, that genre of industry is, Host your mess for less," is basically- all it really was, right? So- Right y- you know, there is a span of control that, that is given up with that. But if you do do it, you're right.
Focus on the bigger picture. Focus on the data. Focus on, you know, where the puck is going as opposed to where you're trying to restore the puck all the time, right? So. 100%. That, very good analogy.The cloud is just someone else's computer Nothing but, man. It, there, there's- Yeah nothing magic about it.
I w- I wish it was, but it's not. So there is a question that came in that I hoped would've been the summary this, but we're gonna- go ahead and jump into it. I, we're totally off whatever your agenda was anyway. I apologize. I knew this would happen.
This one's hilarious. Organization's relatively new to Pure, has historically been managed exclusively by the CIS admin team. How can the CSI admins begin to involve the DBA team and enable them to leverage Pure's capab- Kumbaya. Right Are we going to be the linchpin that has the DBAs and the CIS admins hug it out?
I- i- is it possible for this to happen? Because, man, that is the age-old question, right? And it's been there forever. So let's talk a little bit about, you know, what things we've brought to bear that, that give a modern-day DBA the ability to sit closer to their CSI admin counterparts.
So this one might You guys, this might not be top of mind for you, but it is Yeah. Right? Yeah. Yeah. So one of the hardest things that we've do over the years as DBAs and storage coordinate outages, downtimes, coordinate migrations between devices, right? So NDU is a non-disruptive upgrade, meaning I can upgrade the array, on its
software, I can upgrade the array's hardware, I can even migrate to a new array and never touch the application, right? And that is so incredibly powerful to a database team because they're most systems are expected to be 24/7. Yep. And this is an opposing force to that, where
we had these legacy storage platforms that required these operations. And so the way that I talk to DBAs and storage folks is I can quite literally decouple storage ops and database ops now, like full stop. That's probably gonna make y'all the happiest. Then it's all the cool stuff that we've been talking about, STaaS, clone, deduplication,
performance, all that stuff. So, Ryan, you wanna jump in on that? No, you, you, you nailed it, right? It's, it's The worst thing to ask your business for is, is a downtime, right? Yeah. And that goes, that goes away. That goes away with NDUs.
It's just, it- getting your weekends back and not having to worry about a storage downtime affecting your database, or not having to worry about a migration. Migrations are hard, right? They're not overnight projects. Very. They're- Yeah you gotta do 'em every three to five years on, on legacy storage.
That, that all goes away with Pure, and it's, it's just a, it's a beautiful thing. Yeah, and I've, I've always been a fan of the non-disruptive upgrade from the ability to change the tires on the car as it's driving, if you will. And the more mind-blowing part about it is it's not just controllers and DFM and power supplies and interfaces.
We actually- Right when we were, went from a SAS backplane to an NVMe backplane, that was also a non-disruptive thing. So we, we changed the veins in the body basically while the patient was still alive. That's, that's the level of what NDU goes to, the non-disruptive upgrade. And to your point about the business not going down, I think one of my favorite
think we moved four FlashArray//X to XLs that were running Epic, right? Yeah. And there was no outage. Like, and my wife's a nurse, so when I explained that to her, she's like, "They didn't have to paper chart for days?" I was like, "No, it just-" Yeah. Right you know? So y- yeah, that's, that's a heck of a way to
start, in my mind, with what you're talking about. What, so what other things would, would you throw into the mix that would bring this kumbaya together? I just wanna say, Don, I love that analogy, right? I absolutely love that analogy 'cause it makes it real.
It makes it real and, and a ho- nothing more important than up- Epic uptime in hospitals, right? Nothing but. Nothing but. It could save lives. If you watch I, I don't know if you're watching The Pit on HBO right now, but right now their systems are under ransomware protection, and they're doing everything via
paper charts, and it's chaos. So amplify that times 10 when your Epic is down, and that's- Yep really what you can see. You can see it tangibly, quite honestly. Yep. So Anthony, what else? Or Ryan- Yeah what else can we do for CIS admins and DBAs to kumbaya?
You know, we talked about how great the CSI admins can make the DBAs feel better with NDUs. What, what else is out there that we can kick around? So I think we talked a bunch about snapshots and clones earlier. When I talk to customers about access to data, right? If you're moving a ton of data between system A and system B with backups and
transactional replication, bulk loads, quite literally you can stop doing that. Snap clone, consume the data. Right? Now, all the, all the, security and compliance things still exist, that's part of the conversation, but you're just moving the data in a new way now.
And the idea is I can move that data between systems on the same array instantly, between arrays nearly instantly, as we described earlier. And so getting access to data is probably the other biggest thing that we can do. I look a lot at arrays that are performing, let's just say, they're getting beat up pretty hard.
A lot of times it's, it's just that pattern. A DBA is blowing a backup out to disk just to restore it back into the same array. 10 TiB, right? Like, why waste all the compute and storage cycles to do that? Snap clone, get data instantly.
And so that's another powerful way to really make things a little easier, to go have the kumbaya moment and watch your array be super happy through, the evening when you're not blowing out 10 TiB just to restore it again, right? Do we, do we have tools out there that enable a DBA to manage into their own snaps? You know, 'cause obviously we wanna remove the friction between the DBA- Yeah and the CIS
admin, right? So the way you do that is either through automation or delegation of- Yeah some form of function. I- is there anything out there that we do that offers that?Sure. I'm a huge fan of PowerShell, so I implement all of those workflows in PowerShell, which you can fold right into SQL Server Agent, Ansible, whatever it is you use it from an
automation platform. But our, all of our platform talks REST, so you don't have to use PowerShell, you can use REST API. Ryan, are you in the Python and Bash universe- I'm a Python guy in your Yeah, so- I'm a Python guy, for sure. Yeah. So the back end to all of that is gonna be REST.
So really it's, it's about choose your weapon and how you wanna integrate into your current automation stack, right? So the, so the secret to it then, is something that we already thought about when we designed our entire Purity ecosystem is the API is everything, right? I mean, w- I've always said we're API first when we design. Right.
That we put a function in API before it even shows up in the GUI- Mm-hmm which means we're very serious about the ability for people outside of the storage admins to be able to do storage-related things to their function that matter without bugging the storage admin person, right? That's the whole nature of APIs, right? 100%. 100%.
Hey, jinx. Jinx. Jinx again. Nice. Do we wanna hop, do we wanna hop into the survey questions, Don? I think we have about nine minutes left here. Yeah, go ahead. Throw some survey questions out 'cause we've
got a good flow going here. Let's see, you know, what, what else people are thinking about. Yeah. We want to kind of get a feel for what y'all are thinking. So Nicole, do you wanna pop those questions, I think, or is that for me to go?
So one of the things that we talk a lot about AI, right? Because it's required by, by law, right, for us to talk about AI. Not even just as Everpure employees, but just by market, demands. And we wanna get a sense for where you're going in your Enterprise Data Cloud. So if it's, you know, SQL Server Oracle, what are your plans to upgrade your environment?
Is, is it, are you looking at it at all? Is it something that you're looking at a year and a half from now? Are you, are you never gonna upgrade? What are your thoughts around that? So go ahead and, and, and mash that button, like and subscribe, and tell us what you're thinking with your Enterprise Data Estates.
I'm glad we mentioned SQL and Oracle. I think we've only talked about SAP so far, right? I know. Well, I think we've drifted a little bit Oracle, but Just how I planned it. That's right. Yeah.
You're the mastermind behind it all. That's right. Ask me everything about databases and SAP. And SAP. That's right. And SAP. That's what you had on your slide, which we totally didn't get to, which I apologize. Ah.
But- All good digging, digging right into it. But that wasn't the slide. Yeah. Okay. Do I get to vote here? Ooh, we just had a really good- Or we could just keep, we can just keep going and Yeah, keep going, 'cause we just got a really cool question that I, I do want to bring up later
on, but let's get through the polls and then we'll talk about it. Okay. All right. Do we want to roll up that one, Nicole, and move on to the next one? There you go. Cool. Let's see what we got here. Where are you running your mission-critical
databases today? Let's just say mission-critical databases, including SAP HANA- Yay if you, if you want to consider as a database. Is it on-prem? Mostly, mostly on-prem hybrid. That's what we're seeing most folks go is, is in the, the hybrid story.
Mm-hmm. Especially for mission-critical systems, a lot of folks ran into a cloud and went, "Ouch, that hurts," for whatever reason, money, performance, SLAs, and now they're figuring out kind of the, the balance of where they wanna put their, these mission-critical workloads, right? Because if you're expecting, like Ryan said, 120, 130 microseconds out of an, an array
consistently, that's kinda hard to do in a shared, a shared environment, right? So- 100%, yep. Yeah, and that- Yeah, which- plays into what we just talked about earlier with basis everything else, span of control- Yeah? You, you give up some of that span of control of the variables that matter have control over in an on-prem situation- Mm-hmm that you've kind of wholesaled it to
somebody else. So that when you talk about them coming back, chances are good that it was a lack of control combined with maybe economics Right that, that, that caused them to come back. It's a pendulum, right? The pendulum shifted, I'm cloud only, and it's kind of shifting back and it's settling now on, on let's put stuff in the cloud that belongs
in the cloud, and let's keep stuff on-prem that belongs on-prem. Yeah. I think one of the, one of the toughest things that I wrestled with when, cloud started its rise in 2009, 2010, was I don't even know how the thread got out there, but every single decision maker I would present to was like, "Well, it's just gonna be cheaper." And I'm like, "It's
not going to be cheaper. You don't get rich paying rent to somebody else's apartment. That's not how this works," right? But- Right man, that, that just went viral and it never got, it never got released until the b- the bills started coming in, I guess.
Yeah. There's some, there's some big, beefy SKUs in certain clouds, and you, you run the numbers, they all look expensive maybe you look at 12 months and then, oh, and then you have to have a replica of the same size. Mm-hmm. The next thing you know, you have a $200,000 bill to support just one database, right? Yep.
Yep. Totally. Mushrooms out. Yeah. Okay. Let's go ahead and close that one and pop on to the next one, please. All right. Do we wanna grab the AC ADR question real fast? Yeah. So let's talk about this.
They, they wanna talk about another webinar on Active DR- Mm-hmm ActiveCluster, planned, but let's talk a little bit about something that's on the horizon and how it might affect databases, because we have talked about data mobility, security and data mobility. It's the, it's the three-site replication thing that we have, right?
ActiveCluster to Active DR. Oh, yeah. It's more about the DORA compliance in Europe, but I think there's val- there's validity outside of that compliance. Talk a little bit about how you see data protection evolving with us being an ActiveCluster pair and also have- an outlet for it to Active DR, if
sense.Yeah, I, I think it's gonna You know, I always have these cyber resilient conversations and I say, "Look, we have the tools designed to your SLA, your business SLAs. This is just gonna be one more thing that gets you under that SLA." Mm-hmm. That customer I talked about earlier who doing that two-site ActiveCluster with, HANA replication in between, they're also doing snapshot replication to a, to a third
site, right? Mm-hmm. Which is great. They do it every five minutes. They get their data over there, and they have a catalog of snapshots. Now, with the ability to do Active DR off that ActiveCluster pair to that third site, no longer is it every five minutes, it's continuous, right? Right.
The data's just moving over there at its own pace. It's not waiting for an acknowledgement. And the beauty is, I love ActiveCluster DR. I wrote a blog on this. Active DR makes DR testing fun again. It, it really does. DR testing, as someone who was in IT for a
long time, DR testing was the worst possible weekend of the year. Everyone's grumpy, it's a Saturday- Yeah you're taking downtime, the business is there, the coffee is awful. You're, you do your IT stuff, the business does their stuff, we do our IT stuff, everyone goes home, we don't see each other till Monday.
Now, with ActiveCluster DR, you can do that on a Wednesday afternoon if you want, right? Because you're not interrupting that replication cycle. You're, you're continuously replicating behind the scenes, and go ahead and test your DR anytime you want, and give your business confidence that your DR site will come up in the event of a disaster.
Yeah, it's one of the things I put in that blog on Enterprise Ready too, or it might have been another blog, like, a couple months ago, where y- you know, the declaration of legacy storage is it's six nines, it's always up, it'll never go down. And it's like, all right, I get that, but what's the quality of the recovery when it happens, right? Like, you can't tell me Again, it's one of
these things, you keep backing up all day long, but when you go to recover, you your act together, right? Mm-hmm. And I think what you just said, Ryan, perfectly encapsulates that into, when it happens, we've got a better experience. It's not preventing it, you know, we do too, but it's the follow-through, right?
Yep. Yep. And ac- and DR testing's something that's always skipped, right? Absolutely. It, it's just hard to do, and no one wants to do it, but, you know, it, it Being able to do it live during a, during a business day forms the trust, right, between IT and that doesn't always, always show up.
Yep, and it's a more robust layer of indemnity too, I think. You know, it's, it's not just a checkbox, it's a demonstrable checkbox, if you, if you will. Right. Right. No, I'm excited. I'm excited for this third site replication. I, I have a number of customers that are, that have been asking for it, and I'm excited to
design it out with them too. Awesome. Well, do you wanna close that poll out, Anthony? Sure thing. Yeah. So one poll question here is, where are you sticking Oh, yeah, that's good. And then we'll grab the last one.
We'll call that, and we'll call it there on the polling. Yeah, perfect. Cool. So basically, what's kind of keeping you from upgrading or migrating your Like, what's the pain point? Is it application testing?
Is it storage? Hopefully not, if it's us. You know, licensing cost concerns. You know, lack of time, that's, that's always a problem in enterprise IT, right? So go ahead and, and, and answer that poll. And then I guess, Don, whenever you're ready, we can, in a minute we'll close it out, and
we'll close, close out the webinar, right? Yeah, man. Again, by the way, Anthony and Ryan, I really appreciate your time. This was great insight. I, I like the fact that we brought the conversation into business outcomes versus just, we're really fast and we did this to Purity to make block level performance better
and blah, blah, blah. No, that, that's all cool and everything, but everything that we covered today was a much bigger practical outcome-oriented conversation, right? A- and to your poll question here, Anthony, I, I, I guess, is a database throwing a CD in and hitting next, next in the wizard?
Does it usually Is it more complicated than that? Oh, geez. Yeah. I think, to bring up, like, we're running out of time, but one of the things I had to do once in my career is I had to change a data type, and one data type was a money column in a business system that supported a hospital.
It took me six weeks to change that data type. Oof. Right? Because then the testing cycle to go sure I need to upgrade a system, right? This was an application-level upgrade, but not a database-level upgrade. But same concept.
You're gonna You change a thing, you're gonna test the heck out of it before you move forward. Mm-hmm, mm-hmm. And so with Pure, right, what, what was The slowdown was every time I made a change, I ran the script to upgrade, I found the bug, then I had to restore the multiterabyte database, which took two days. I'd go do something else for two days.
I'd come back to the problem, I'd run the script. Well, I'd get back in Kontxtual, put back in the front buffers, get into the, on the application, run the script, find the bug, go restore it for two days, come back, right? On a FlashArray, snap clone, run script.
Oh, s- we're broken? Snap clone, run script, right? So just think about the acceleration of time to get the upgrade done when you have When you can decouple space, like the size of the data, from the things that you need to do with the data. And I think- Correct the key to that, too, is
if you were the DBA doing this, the right API integrations would allow you to do this independently without bugging the storage administrators all the time too, right? Right. 100%. Perfect. Well, that's a heck of a anecdote to, to finish us out here- obviously, with business outcomes as the main themes.
Again, before we head out, you know, Ryan, you got any parting shots you wanna throw out there? Yeah, I'm wondering why Anthony had so many bugs in his code. Yeah. Poor syntax. Ouch. Hey, I'm, I'm Pure at- He didn't mean, he didn't mean those kind of shots dude, I'm Pure
at syntax too. Oh. Oh, sorry. I, I hate syntax, so I, I grieve for Anthony if he was going through that just, I, I appreciate everyone's time today. You know- Yeah we really honestly care about databases and your mission critical databases. I think it shows, me and Anthony being here and, and, and us being, you know, just part of
this database team. And, you know, keep it coming. Like, let's, let's continue the Anthony and I are always available, and I think we may even have a plug for a, a conference coming up here in a second too, right, Don? Yeah, absolutely. Nicole, throw the slide up.
I wanna talk about that real quickly. And by the way, there's a parting shot from somebody who asked a question earlier. Storing a sensitive DB in the cloud while relying solely on the provider's trust is like keeping cash in an apartment building mailbox." That's a heck of a metaphor. That is a heck of a metaphor.
So- That's gonna become a laptop sticker. Yeah, right? Exactly. I like that idea. So anybody, if you're still on, one plug I wanted to throw in here at the end is Pure//Accelerate is already scheduled and happening. We're going through the machinations behind
the scenes to plan for it to be a pretty epic thing. It's gonna be June 16th to the 18th in Las Vegas again at Resorts World. And if you are interested in attending, definitely register, reach out to your account team, talk about the logistical details of it. The good thing is, is that Ryan and Anthony will both be there, so you can ask them all
your DBA questions and confirm that they're not chatbots. I'm gonna be there as well, so you can confirm that I'm not a chatbot as well. But definitely make some time if you can to get to Las Vegas and see us June 16th through the 18th. And with three minutes over, which is actually very, very good, Anthony, Ryan, thank you very much for joining us today.
I really, really enjoyed this conversation that I told you was gonna life of its own. And, uh- Mm-hmm I think it worked out well, just like it has with every other Ask Us Everything. And on that note to everybody, remember, Ask Us Everything is about to amp itself up to every two weeks. So keep an eye on what the invitations look
like, and also remember to participate in the Everpure community that's out there to be able to interface with people like Ryan and Anthony who are in there all the time. So again, thank you very much for joining us on Ask Us Everything. Thank you to Ryan and Anthony. And I'm Don Forbin.
I'll see you on the next Ask Us Everything. Take care.
Anthony Nocentino
Senior Principal Field Solutions Architect, Everpure
Ryan Arsenault
Principal Field Solutions Architect, Everpure
Don Poorman
Sr. Technical Evangelist, Everpure
Got questions about running databases at scale? Get answers.
This month we’re diving into databases—and how Everpure (formerly Pure Storage) helps infrastructure teams support modern, mission-critical workloads with less risk and operational overhead.
You’ll get an overview of how our next-generation storage platform enables:
- Consistent performance and scale for AI-enabled database workloads
- Simpler, automated operations with Everpure Fusion™ presets and workload setups
- Built-in cyber resilience and high availability
- Predictable costs with non-disruptive upgrades
We Also Recommend...
Personalize for Me
My Updates
No updates at this time.