Dismiss
Innovation
A platform built for AI
Unified, automated, and ready to turn data into intelligence.
Dismiss
June 16-18, Las Vegas
Pure//Accelerate® 2026
Discover how to unlock the true value of your data.
Dismiss
NVIDIA GTC San Jose 2026
Experience the Everpure difference at GTC
March 16-19 | Booth #935
San Jose McEnery Convention Center
45:42 Webinar
The Modern Virtualisation Journey at SiriusXM
Learn how SiriusXM modernised its infrastructure—key technology choices, migration tips, and strategies for scalable virtualisation
This webinar first aired on 18 June 2025
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
Welcome in. Super excited to be here for this session. Uh, my name is Adam Sweidler. I am director of product marketing for Port Works by Pure Storage, and I'm joined by Nate Mason, director of platform operations at Sirius XM. Before we get started, I wanted to just ask the audience members, how many of you are Sirius XM subscribers?
Any serious? Yeah, the majority of the house. Awesome. Thank you. I too am a Sirius XM subscriber both in the car and on the app and uh on behalf of all listeners, Nate, I wanna say thank you for the wonderful experience that you're creating for us out there on the airwaves. Thank you, it's a joy.
Why don't you tell us a little bit about yourself and what you do at Sirius? So at Sirius XM, I, you know, handle a lot of the virtualization and cloud native, uh, containerization, basically anything, you know, relating to. You know, Kubernetes that we, we have a Hay corp stack as well, uh, and the Pandora Cloud as well as, um, um, yeah, basically virtualization,
which is a hot topic, you know, these last few years, right, so, um, you know, I, I, I, I mentioned this before, but we do like to like diversify our platforms so that we can shift when needed to like the industry pushes a certain direction like it has, you know, we, we can kind of adapt to that so. That's really what it is. It's about just keeping the right technology on
the floor to be able to service, you know, our workloads. Yeah, and we're gonna come and talk in more detail about virtualization in particular because, hey, that's in the title of our session here, but I wanted to maybe just, you know, take a step up and ask you like what were some of the fundamental things that caused you to look for, uh, a more agile and dynamic environment.
Uh, I mean, you know, we've got, you know, developers, we've got, you know, campaigns that launch, we've got different things that happen, you know, throughout the year towards the end of the year we have, you know, keep up with our OEMs, you know, they're, they're manufacturing and, you know, how fast they're rolling cars off. So, you know,
we gotta adapt quick, especially with our partners like, you know, Amazon and things like that, you know, with Alexa, so. Uh, that's really you gotta move pretty quick, so we, you know, we wanna make everything as self-service as possible and, you know, just not really have a lot of red tape when somebody needs a resource to get
something done, right? Like, you know, we have really smart people and to not get in their way is a wonderful thing. So that's really it, right? Move as fast as you can and you know, let's, let's support that in any way possible, right? And in some of our other conversations you
talked about like getting out of the trenches in a sense and and building a platform to help you out. Can you elaborate a little bit about what the trenches looked like? Uh, yeah, I mean like we had pretty tall verticals, right, um, multiple groups that they overlapped in some areas, uh, where it's either, you know, working with servers in the actual data centre to all the
way up into the cloud offerings and, you know, maintaining images for, you know, AWS or like wherever your cloud provider is actually we use all of them, right? So you know it it it you end up with this really, you know, one group managing all these things from like, you know, bare metal like hardware all the way out to like cloud stuff you know.
You gotta kind of flip sideways at times, right, like get, get the pieces horizontal, you know, remove a lot of the stacks, you know that other people are overlapping each other with and then just kind of, you know, break it down from there and say, alright, well, you know, now we have a data centre operations now we have a systems engineering group that
can handle, you know, kernel level type stuff operating system, you know, now we have, you know, platform, you know, my team platform operations now we're looking at, you know, virtualization and containerization. And then you know one stack above that, you know, platform enablement to really push it, you know, and evangelise the developers say hey this is what you're doing it's really cool,
you know, it's gonna be quick, you know, get on board, you know, and that's pretty much it. So and those are the developers creating those experiences for the listeners out there and and enhancing those, uh, those experiences with new features and and new content as well for sure, um, so let's take a step back and and maybe talk a little bit about how Pure was first
introduced to Sirius XM. you talk a little bit about your early exposure to Pure? Yeah, you know, we were coming off of a, you know, a legacy platform, so, you know, technology had changed from what we were had been running for, you know, 10 years, let's say, to, you know, uh, now with pure it's, you know, it's a technology shift,
right? We're, we're looking at a smaller footprint, you know, NV RAM, a lot of flash, right, versus, you know, rows of just, you know. Discs, right? So for us, you know, if you're gonna change, you might as well do it right.
If you're gonna lift and shift, like do it like just get rid of everything and you know, get that smaller kinda footprint, but like, you know, our database folks, some developers like they know latency. Hey, I'm off by a few milliseconds what's going on, you know, it's like, so the observability was a huge piece as well.
So being able to give them a, you know, a a console, say, hey, look, if you, if you're having issues, you can look at it live, you know, so. Observability, smaller footprint, you know, fast, multiple tiers, it, it pretty much hit it, so. Um, and so, you know, maybe compare and
contrast a little bit. You did this a bit, but compare and contrast a little bit what it was like in the sort of pre-pure days to what it's like now. Um, I mean it's faster. I mean there's that right, which is huge, but uh observability is huge. I mean that's really was a big piece to be able to, you know,
let folks look at their, you know, workloads live and not have to kind of, you know, interact and say, hey, we got a problem. What's going on? Alright, well, we'll have somebody look at it. Now you know we engage like when something's actually, you know, if there's an issue you know we need to bump them up to a different tier or something like
that, you know, like faster like a X90 versus like C70 or something like that so you know we can move the workloads around before it was kind of static. You kind of like, hey, look, this is essentially the, the fastest this system will run and that's it like you don't really, there's not much room to move there so.
You know, different tiered storage is huge for different workloads, whether it be virtualization, containerization, database, you know, web apps, you know, things like that. So, um, that's a huge help to be able to, you know, just to, you know, have different, you know, speeds that you can run,
so. Um, so I wanted to mention if any of you have a question as we're going through, uh, the conversation here, feel free to raise your hand. We've got a nice small intimate group here. We're happy to take questions live if you want to hear more specifically about some aspect of the journey that Nate has been leading Sirius XM on,
um, and we'll keep going in in more detail here, um. The other thing I'm gonna do, I'm gonna give you a heads up, we're gonna have some audience participation. I've got a couple of polls coming up where we're gonna kind of get a pulse on where you all are in the audience. So if you don't mind, get ready, get your phone out, there's gonna be a QR code and a question
that that that we're gonna ask you coming up here. So I'm just preparing you for that, um, but maybe you could tell me a little bit about how the platform architecture like translates into better listener experiences and, and like sort of developer innovation and productivity. I mean, it's that, right? It, it leads to, you know,
the partners are everything, right, whether they're rental car agencies, whether, you know, again, Alexa, you know, with Amazon or, you know, partners of the OEMs with the manufacturers. So. You know, when you got your C-suite making deals coming up with cool ideas of, you know, how to do things they don't wanna hear it's,
well, we're gonna have to respond. We need a year like no like we wanna do this like next month or something like that. So I mean that's really the driver and you know, essentially when they when they come up with a campaign or they come up with an idea and come out of it, you know, I'll call the guys at Amaz hey what are you guys doing? Hey, we're gonna do this. Yeah, we kind of did that last month we're
gonna do the same thing again so we just, you know, it's really how fast you can react and you know, put things on the market, so. Um, you know, it gets easier over time when, when you, when you have a flexible platform that you can operate on, right? So it's not, you're not really stuck to like a
rigid way of doing things that, you know, you can kind of start a clock in the clock like alright this is gonna take us a long time to do now it's like pretty quick, right so. Um, and can you tell us a little bit about how, uh, you empower the developers and sort of the philosophy that you have around developers accessing the platform. I mean, we really just don't want to get in the
way and I don't wanna tell somebody no, right? Like, cause they know what they do, they're really smart people. We've all, you know. Economic, you know, you kind of shrink a little bit. You expand depending on, you know, how the economy's going like that. So the people we, we have are really smart, so I, I don't wanna second guess them,
you know, if they know what they wanna do, you just wanna get them running as fast as possible. Just here, go do your thing, you know, uh, uh, you know, self-service, don't worry about budgets, we'll take care of it, you know, just here's a, a, a, a massive platform that you can operate on. So that's really what empowers them, you know.
When they don't have to ask for something that's wonderful, right? It allows them to just proceed at their pace of innovation, not dependent on an infrastructure team to provision storage or anything like that you don't have to tend meetings on, you know, capacity and this and that and you know how we're gonna operate it like here it is, go for it.
So, uh, and then some of the conversations that we had previously you talked about this sort of ethos of if you build it, you run it, and that's how you, you sort of share it with the application teams you wanna talk a little bit more about what that means to them? I mean nobody's gonna know it better than the developers or whoever is, you know, creating this app or you know what they're doing so for
sure I mean they have to run it like operationally but you know it also puts them in a good mindset too of like from A to Z, like really what are we trying to do? We're not gonna be able to just hand off the operations to somebody else. We can't write bad code we can't, you know, outages and things like that so you know when they know they own it.
It gets built correctly, you know, not, not, not that it wouldn't be, but you know they're thinking about more just than just, you know, tossing the app to an operations team later on, right? It's like they're building in the features and the operational support that they need within the app.
So, um, yeah, again, when you're, you know, that sort of mandate, you don't wanna. You know, you don't wanna, you know, just. You know, make them wait essentially so that's really what it comes down to it's just really quick service. So yeah, in many organisations they are doing exactly what you are doing,
which is really trying to remove all the roadblocks and friction from the developers and the AI teams so that they can really focus on writing code or building models or whatever is and not, you know, be bogged down with infrastructure requests or decisions or like you said 15 meetings and capacity planning and stuff like that, um, and so you all were sort of on a modern application and a platform.
For journey, you know, I think for for a number of years and then along comes this little disruption in the virtualization space thanks to Broadcom and and the the upcharges of the increases in licence fees, um, tell us a little bit about sort of your virtualization journey. Um, I know that you have both VMware and you have some of the other, uh, uh, virtualization in place previously, but why don't you start a little bit about the
history of your virtualization and now how you're thinking about reacting to the disruption that's taking place. Yeah, I mean, so we were diversifying VMware before Broadcom, so, uh, you know, we had Rev, got that in the door, you know, the red hat, yeah, red hat enterprise virtualization.
So you know we, we did what we could, right? So as far as just even, you know, just moving some workloads off, let's, you know, tighten up the bill for things like, you know, maybe dev environments, test environments, you know, keep the production stuff where it is, but so we had that other platform as an alternative.
So when the news came down, like, you know. Sirius XM and you know, we're competitors with. Inside the dashboard, chips, you know, Broadcom is, you know, we, we pay each other licencing fees to get around, you know, it's, it's kind of ridiculous. So when that happened,
like we knew where it was gonna go. C-suite's gonna call, Hey, can you get us off there next month? No, we can't. However, We're halfway in the middle of a 3 year deal. We got a year and a half now we can start diversifying a little bit more, and that's kind of where OpenShift came in the door.
So, um, yeah, and it didn't take us long. We were like, yeah, this is gonna move quick, right? Like I don't wanna, you know, we know when that renegotiation is gonna happen and we wanna be as far out of it as possible, um, because, you know, I'm not really knocking Broadcom but they have a plan. They know exactly what they're doing and
they're doing exactly what they said they were gonna do so. You know, the rest is up to you, right? So yeah, for sure, uh, I imagine a number of you in the audience may be uh looking at, you know, significantly increased, uh, licence fees and, and in some cases you may have even renewed because in a similar situation to what,
uh, Sirius XM and others have found, you know, it's not necessarily an overnight or a couple of weeks or even a month or two to get, you know, off of of that platform and on to something more modern but this is the trigger moment which is causing companies to start building that plan to avoid being caught in that situation. Uh, the next time around in the future, um, yeah, this is not a 10% at 20,
this is a 5 X7X9X like you're completely outside the realm of what you were doing before, so it's, it's a big, you know, we're not talking a few percentage here it's, it's a massive shift. So I was just talking to a guy that, you know, ran 70,000 VMs for a bank and he's like we, we could, we can't do it that fast, right? So, you know.
Little bit of head start would help, but. Um, so one thing I wanted to do now is see if we could get a little bit of audience participation. If you can pull out your phone and scan that QR code, we're trying to get a sense of where the people in this room are in their Kubernetti's maturity journey.
So if you all scan that and then answer uh one of those questions, are you kind of just getting started? Are you, uh, sort of in this intermediate phase where you're doing test and dev or are you at the point where you do have some production workloads or are you, um, you know, sort of pretty committed to committed Kubernetes and running many,
uh, production workloads? And so we see those results coming in and it looks like the majority of the room, well, we got some. Some bounces coming. We've got a good distribution actually showing here. With the majority of folks kind of in the earlier stages of their their Kuberneti's
journey and that's consistent with what we see overall in the pure customer base and even in the overall market. Kubernetti's, you know, has been around for about 10 years, but it probably only started to pick up steam in the last 34 years, maybe 5 years in terms of uh enterprise data centre platform to,
to, uh, run applications. But now with the disruption that's happening as a result of Broadcom and the ability to run. Uh, uh, virtual machine workloads inside a container without having to rewrite or refactor the majority of that code that's proving to be an attractive path, uh, for people and what they might do with some or all of their their VMware workloads,
um. Maybe talk a little bit about. Why Red Hat was such an important partner for you all in this modernization of virtualization. So you know when we merged with Pandora, Pandora had a massive, you know, an internal cloud that we used based on the Haystack, right? So we use both and we're fine with it,
right? But you know, OpenShift really when Cuer popped off like in 2018, that's when you, you, you really start to look like, all right, this is gonna change a lot, right? This is now we're talking about hyper converged containers and virtual, you know, virtual machines like essentially running. You know, together, so it's like, alright, this is a kind of a consolidation of,
you know, 3 or 4 different platforms down into, uh, you know, one, but you know I, I would assume like a lot of, uh, organisations like ourselves at the time, you know, had lots of VMs that were running lots of containers. Well, you know, it'd be cool to just get rid of that VM layer right? and then just kind of move things over to be
native, you know, container based. The apps are kind of written like that already. You just have to move the operational part of it around a little bit and. You know, now you're essentially and that kind of also sets you up to put things out to your different cloud providers as well, right? So you know it, it's kind of the right step when you see a technology that is,
is kind of heading in the right direction, you could, you know, you wanna get on board as soon as possible. So and you have, uh, in your, in your sort of legacy VM estate you have both VMware and kind of the older red hat. Virtualization uh capability so talk a little bit about how uh the OpenShift virtualization,
the more modern one is providing a landing place for both of those types of VM workloads. Um, I mean, it's still KVM under the hood, right? So, you know, whether it be, you know, red hat virtualization or Qbert or, you know, OpenShift, I mean, it's still kind of the same thing, right? It's, it, it's like the hypervisor being
wrapped in like a container, you know, kind of, you know. It's really the same thing on the hood, so I mean we were pretty comfortable with it, but you know for the most part uh it gives you more flexibility when you're not dealing with like multiple disparate kind of platforms, right? You can kind of send it out, you know, capacity planning is a little bit better.
Hey, we need to add 1020 more nodes, you know, whatever it may be. So you know it get you do get consolidation out of it which helps, right, rather than, you know, multiple control planes, you know, from all these different platforms you can kind of really start to consolidate down. Uh, which, you know, again, where the storage comes in as well,
right? That's, that's huge, being able to get, you know, low latency, you know, when you start, you know, a hyper converging, you're, you're. Again, I, I, I mentioned this before, you're making your hardware really work for you, so there's not a lot that's just sitting around doing nothing,
so, you know, it heats up and. You just keep adding to it, but that, that's really a nice common platform to be able to operate all of it, so. And again, if you have any questions, raise a hand and I will spot you and bring you into the conversation. I back there.
Or I'll just call on you, um, talk a little bit about the migration process for these VM workloads and you know sort of how you approached it and what you learned through going through that process, um. It's a migration. It's, it's not difficult, right?
So sure you may have to take some outage right for an application, you know, or, or move workloads to different data centres where you're doing it, but I mean it's not that it's not a huge complication, to be honest. Uh, I think the tools that they provide, you know, whether it be a native Kubernetes or like a redhead open shift type product, um, you know, it's, it's well rounded.
They know what they're doing. They understand like, you know, VMs, we've been made to migrate VMs for years, right? So it's, it's not, it's not completely alien, right? So, but the migration is not too difficult, right? As long as you have a solid storage subsystem,
you know, the data can move quickly and you know, you can do 300 VMs in a night, you know, and just schedule it out over some months and kind of clear out that platform, so. It's, it's not And um we were talking before you know sort of you approached it in phases uh of how you you migrated talk a little bit about those phases as as an example perhaps or a best practise that folks in the room could maybe
learn from. Well, you wanna be comfortable with it, right? So you know, sure, start like maybe in a development environment, you know, and then work up to like test and QA so you can really make sure the performance is there and then you start moving into production. So you know it's an iteration. It's not like on day one,
you know, you haven't, you know. A platform you're not super familiar with and you're not you you're comfortable with it but you're like you're not sure what's gonna like it's not like that by the time you get to real production workloads you're you you've been through the gamut you you know how it's, you know, you get your belts and suspenders, you're monitoring,
you know, monitoring your operations team, like, you know. Is comfortable with at that point at that point you're like, sure, let's just get this stuff moved basically, right? So makes sense uh and then can you talk a little bit about the role that Port works? Oh, we have a question. Yeah, there are workloads that you can't,
uh, you know, modernise or have a lot of difficulty, you know, getting onto onto Openhi from? So the question is, are there workloads that are, you have difficulty in getting on to OpenShift and migration? I think like appliances right from certain vendors, whether they be, you know, uh, a networking appliance or something like that,
sure they have to support it, right? And, and most of them do, so you know, a few things might lag behind, you know, until that vendor can catch up and, you know, rewrite it for like, you know, OVA or QA or you know whatever type of image there, but, um, other than that, no, not at all it's really just support like there's nothing that we wouldn't put on it. You know, it's, it runs pretty well,
so. And as you were investigating the alternatives, can you talk a little bit about other things that you looked at and why they were maybe less attractive than the uh Red Hat OpenShift virtualization solution? Sure. I mean, it's got to be enterprise grade, right? We may have, you know, an office in Ireland
that, you know, it's like a telco closet where you know you can put together like Newtonics is a good product, right? It's, it's a really good product, something like that it's fits perfectly for, but when you're, you know, consolidating like data centre, it's gotta be a little bit larger, you know, kind of scale, right? So, um.
Yeah, that's really what comes down to and cost is a thing as well, right? So. Um Yeah, I mean, we actually use a little bit of everything to be honest, uh, it's not, it's not one or the other we're not like that, right? We, we don't just put all the eggs in one basket and go for it like if, if there's an opportunity for,
you know, a certain product to shine, like we'll use it absolutely, but we don't mind, you know, managing multiple platforms as well. So, so, um, did you look at Newtonics as a potential large scale replacement for VMware or you already knew the place where Newtanix played well and. We do satellite sites, you know, studios and we're like,
you know, this will probably work really well here, but again, you know, big large scale kind of data centre consolidation. I mean it was really kind of one direction we were going and you know, Cert and OpenShift, I mean that's gonna evolve over years as well. I mean, VMware has been doing this for 30 years, right?
25, 30 years. I mean they've got, they've got a long track record. in this, right? Cuba, it's been around since, I think accepted by CNTF in what 2018. So I mean there's certain things, you know, hot adding CPU and memory and disc and stuff like that they're gonna catch up with.
But you know it's on the road map and it's being embraced, so it's, it's gonna happen. It's just a matter of time and, and, and if you can live without those certain features for now, like it just gets better in the future, so yeah. Um, and so talk a little bit about how Port Works in particular came in and address some of
those sort of gaps that exist with with Cert and OpenShift virtualization without Port Works. Yeah, I mean Portworks is, you know, you need a solid CSI, right? I mean you have the storage is, is everything for virtualization, um, even containerization as well, but like that's really what it.
Yeah, I mean, Portworks allowed us to do multiple tiered storage, you know, again with different, you know, like X90 C70, so we can have T1, 22, T3, you know, backup we just roll dump space. We can put stuff out, you know, as much data as you can dump out there essentially. But you know, port works allows you to do that, right?
It's, it's nimble enough to have a multi kind of tiered system and, you know, you can change things on the fly, which is huge. You don't really know until you're migrating, you know, 34, 500 VMs in a night and you're like, we just ran out of space. Like we, we gotta do something, right? And then,
you know, open up the CLI, you know, put works is very well, you know, documented, right, so go and expand the space. It does it on the fly, no adage, no down time, you're like, all right, now we're back up to 50%, we're good. Let's add some more type situation, so.
It's, you know, it's, sure, OpenShift comes with its own technologies built in, depending, you know, what the middle where it may be, but, you know, for us, we wanted something really interactive with our pure storage, you know, and obviously, Portworks is gonna do that, so. Um, you talked a little bit about, uh, multi-cloud environment.
Maybe tell us a little bit more about your current environment and how the virtualization is playing out in that environment. Well, um, virtualization, we keep on prem, uh, multi-cloud, I think is good for, you know, you know, green field, blue field, you know, moving stuff for data centres. Like if somebody,
if GCP has a deal going on for a month, we'll move workloads out there. If AWS has a good deal going up, we'll switch those, you know, so you gotta be nimble with it, but OpenShift allows you to do that, you know, to be able to push, you know, workloads out whenever you need it, whether you need it on prem or you're expecting a spike through the holiday season,
gift cards, things like that, you know, we might leverage multiple cloud providers depending on, you know, what, what regions, what areas we're trying to hit at the time. So that's a cool thing to not be locked in and to be like, hey. You're fluid, right? for sure and, and, uh, Portworks helps provide a bit of that flexibility and agility on on the
persistent data. Uh, side of things, um, in fact, we sort of see virtually every, uh, enterprise going forward managing, you know, their own data centre footprint and on-prem, uh, infrastructure along with at least one and sometimes two public clouds and so now sort of going forward we're gonna be in this situation of having to identify based on,
you know, performance, security, uh, and cost. Uh, the, the right place to run a workload. Historically it wasn't once you put a workload somewhere it wasn't necessarily very easy to get it and run it somewhere else, but in the Kubernetes world and together with Port works we actually do make it pretty easy to move applications from one Kuberneti's cluster to
another regardless of where those clusters are on prem or in the cloud and so a lot of customers also have experienced. success and and um acceleration of speed with the cloud, but the trade off being high cost associated with getting that and so we're we're also hearing a lot of customers like Sirius XM looking to run a lot of those workloads on prem.
Because they can do it more cost effectively now as the virtualization disruption comes in, almost all of those workloads were already on prem and so there's a natural inclination to find an on-prem replacement but also all of the cloud providers offer sort of a hosted version of the VMware stack, and I'm wondering, did you all look at the cloud provider versions of those? Nope. And then why not?
Why would we? Because you knew that the cost would not work out, yeah, so our initial like kind of entry into the cloud was just like, you know, a, a mandate. Hey, we gotta get stuff out of the cloud. So what do they do? They just start spin up, you know, EC2 instances like that could run anywhere.
Why is, why are we paying somebody else's data centre, you know, why wouldn't we just do that on prem? So, you know, for things like that, like, sure, edge is edge, right? There absolutely are workloads. I don't care if it's a VM and EC2s that needs to be out, you know, as close to, you know, the customer as possible.
So sure, I mean, they, they automatically get sent there but you know, back office, you know, enterprise service buses, things like that naturally fit, you know, basically in your own data centre like very well, right? Because that's where most of the traffic's interacting with other applications and things like that.
So, um, again, like we, we put things where they need to be, right? So. And so in the keynote today for those of you that saw it, we talked about in the virtualization section at a high level there are sort of 4 pathways available to customers. The first one is to stay with VMware, negotiate hard, optimise your, your spend there.
For most companies that's the lowest risk option for sure also probably a pretty high cost option even with hard negotiating and things like that. Second option is to maybe change the hypervisors, look at something like a Newtanics or a Hyper-V or you know something else that's out there, hyper converged infrastructure.
Third option might be to go to the cloud in one of the hosted VMware service offerings, uh. Talking a lot about the Azure offering in the in the keynote today and then the fourth path, which is the one that I personally would say is like the the modern path, the path of the future, is to take those VM workloads, wrap them in a container using the ubert open source project,
and then run them without having to refactor them in a lower cost environment. Um, again, your choice of on-prem or, or, uh, in the cloud and so we've seen Sirius XM, we've seen a number of other large customers say, you know what, we believe that our future is more container based than VM based. And that is helping to encourage us to start looking at moving these VM workloads into that
container world using this Qvert, uh, uh, and, uh, open source virtualization, uh, open shift virtualization, sorry, uh, as the, the sort of run plane for these things. So one of the things I wanted to do was actually get a sense again in this room if you wanna get out your phone and, and, uh, audience participation time. Of those four options I'm curious which one if any you all are considering,
um, including the last one which is still evaluating no decision that would include things like we we've re upped with VMware, you know, unhappily but we're doing it because we don't have a choice in the short term but we're building a plan to reduce our dependency so let me get a sense of uh what the audience is doing. So right now, um, it's pretty evenly split and I think that is consistent with a lot of what
we're seeing, uh, in the marketplace, um, for some organisations, you know, they, they may not have any experience with containers. It just, you know, it might not be the right fit for them. They don't have a need for rapid application development, which is part of the. The reason for going to a a containerized and and modernised platform so for some folks
definitely um we're seeing that but then thankfully we're seeing a lot of interest in the open uh open shift virtualization and cubet solution uh in this audience as well and I really do believe that uh red. That along with uh Pure and Port Works have identified this desire in the customer base to have autonomy when it comes to where we run those those workloads and those applications. People are feeling really burned about this
VMware situation and they really want to avoid getting into that situation again, um, and so you know, running with a platform like OpenShift gives you that flexibility to swap out different infrastructure components underneath it, swap out different cloud storage pieces in it. Um, and it does, I think, provide a, a really agile platform,
and Port Works comes in and provides the persistent data component of that, of that kind of agile platform. Um, again, any questions in the audience feel free to, uh, to raise your hands, um. So let's talk a little bit about what's next.
And in particular I know that you're passionate about not having VMs forever in the organisation. Talk about that a little bit. Yeah, that's really where you wanna go, right? Um, VMs are cool and all, but you know it takes more resources, uh, to be able to, it's a lot easier to move a container from one platform to another or a different cloud
provider than it is a VM, right? That's a, that's a pretty big lift, right, to just spin things up naturally. You know, and, and say all right, now we gotta send something from, you know, you know, New York Data centre to like, you know, San Francisco or something like that. I mean that's not gonna happen quick,
right? So ideally you want your containers, you know, to be, you know, globally kind of the, you want them moving everywhere wherever they, you know, they need to communicate with and sometimes other countries, right, depending on manufacturers, you know, like they're all over the world, right, so. Um, that, that's really what it comes down to,
kind of lighten the low a little bit, you know, get, get the, you know, get the containers off of the VMs, get the VMs, you know, they, they should really just be used for appliances, you know, things like that kind of have to run there, right? So, uh, I mean containerization, that's really where you wanna be,
I think for everything, right, as much as possible. I mean we're even looking in database containerization at this point, right? BRM, you know, supply chain, I mean. Everybody else is kind of catching up, right? All of our, you know, software vendors are catching up and saying,
hey, look, this is the right way to go. So even for our streaming, you know, and broadcast, like, sure they're based on VMs at a certain point, like different channels and things like that, but you can see them moving as well. They, the whole industry is moving. It's not just us saying we would like to do this.
It's that's where it's going, so yeah. Um, and so as the industry, uh, moves in that direction, maybe tell us a little bit about what you need from partners like Pure and Port Works to support you. Um, yeah, I mean we need people that, you know, can concentrate on their product and, you know, and, and partner with, say, hey, the way we use it this year,
I would expect would change before next year and we say, hey, Port works, you guys need to do this. We have a workload, we have something crazy that happened and you know take this to your engineering group and you know, come back and sure, you know, let us, you know, let us do that. A couple months later they developed,
you know, a new thing and you know, now we can now integrate it into the storage system, right? So. Um, yeah, I mean, really, we're not in the business of designing storage platforms. That's not our business. We need to concentrate on Audio entertainment, you know, podcast, you know, things like that, so.
That that's where you come in, yeah, um, and tell us a little bit about we've talked a lot about sort of the development teams and people that are creating the actual experience for the listeners. Part of that is also data, data science and AI. Let's talk a little bit about how your platform is also supporting those efforts. Well, yeah, sure, I mean Pandora with the Music Genome Project,
you know, that's what, 20 some years old now, right? So it, it started way back then, right? And now we're getting into much more. I mean, we're talking, you know, it's not just satellites, shooting music to the northern hemisphere of, you know, of the Americas. It's not that anymore. Now we have different chips coming out where we're, you know,
now we're getting two-way communications, uh, you know, telemetry data, manufacturers love it. They were like, oh great, we're gonna be able to put ads on there and it turns out one of the greatest things they were able to do is find out where to build service stations for your Honda, your Lexus, like whatever you have, right?
That's really they started so the data changed everything like what people thought they wanted to do, especially for, you know, the manufacturers. So now it's like they're coming up with crazy ideas and like we gotta process all this stuff, right? So, um, that, yeah, that's it's grown over time for sure, right? AI is a thing.
It's gonna be here, you know, ML, whatever you wanna call it, but it, it's here to stay. So, you know, we have to adapt and be able to give our partners the information they need, so. Yeah, I was just looking at some of the notes from our earlier conversations and I think you talked about having like over a million data points on each customer that you then want to
try to figure out how to best serve that customer and provide content, not just for us again manufacturers as well, you know, they want, they wanna know, I mean a customer's not just one person's customer everybody's customers. So there, there's millions of data points that you know we kind of we have to sift through and say hey listen,
you know what's what's gonna be best thing for you? You know, traditionally in audio entertainment, you know, you just, you have a DJ, you blast music out there and that's it, right? Now it's like what do you wanna listen to? It's probably different than what I wanna listen to at a certain time of day and then
your children wanna listen to something else, right? So you have to adapt and be able to say, all right, you know, this is it, it, it becomes more atomic. The customer becomes uh an individual, not just millions of people listening, right? So and that's where it gets fun. You know, trending, you know, like, you know, what have you been clicking on as far as like
that type of music, you know, we, we, you know, our algorithms right now have, have created stars, right? from some obscure person playing a guitar to now they're, you know, at the CM CMA, you know what I mean? like. So it's a little bit of that as well, right? So we gotta deal with the music industry,
but very atomic as far as the customer goes like we need to hit everybody specifically it's not, you know, just blasted out there anymore. It's, it's hyper personalised. It's very hyper personalised, yeah, because you're doing, you know, Spotify is a thing, right? I mean that's kind of evolved into where people starting to create playlists for each other and that caught on and that's like a really,
it's a really great service, but you know we have to do things in a different way, so. Yeah, um, you mentioned Pandora. Many of you may know that Sirius XM acquired Pandora a number of years ago. You also mentioned the the the Music Genome Project, and I think it may be an answer to a trivia question that like Pandora was actually the like the very first recommendation,
uh, algorithm and engine even before Netflix started doing the, the video recommendations, um, and so that's something obviously that you all are are taking advantage of. Let's talk a little bit about you personally and your career and how this platform has benefited you. I mean, you know, you get to sleep a little bit longer when things aren't breaking,
right? That's cool, uh, career wise, I mean, I, I would expect every organisation, you know, can recognise good talent when they see it if you, if it's not making a really good decision once, you need to build a rapport, right? trust level with, you know, your management leadership and things like that. So over time if you continue making the correct
decisions. You know, the cream rises to the top. So I mean, unless you're just failing at everything, then you know I, I would expect mostations to kind of support you and say, hey, what do you need? Oh you need a bigger team, you know, you need this, like, you know, budgetary, is everything OK?
You need more hardware. So when you're doing the right thing, you just get a lot of support from everywhere. So it's gonna, good things will happen for sure. And what we, what I think you've experienced, we've seen it with some other customers as well is you have a sort of small but high powered team supporting the platform.
And the platform that in turn supports hundreds or even thousands of developers if you extend out to the partner ecosystem, right? And, and that sort of ratio of platform team to development teams is, is, is quite small. There's a very small platform team supporting, you know, many, many thousands of, uh. And this gets back to the whole self-service. Let's not get in your way because we don't have time, right?
I mean, there's just a few people managing. You know, a platform that like thousands upon thousands use, so. Yeah, you gotta, you gotta do what you can operationally, right, or else you're gonna sink, right, and nobody's gonna give you like, you know, increase your staff 10 times, 10 times, you know,
so yeah, you gotta kind of work out your own problems and, and you're a very humble guy, uh, um, because I think maybe last year when you were here your title was a little bit different and this year you are the. Director of platform operations um and so you know I, I can I, can I get you to talk about how that's this helped get your your promotion and get you
just right, make the right decisions that's it. I mean, I think, you know, I wouldn't want to see anybody in an organisation that doesn't get kind of rewarded for a lot of hard work and doing what you're doing, right? But again, it's a trust thing, right? I could have chosen like 30 different platforms to go to like when you do the right thing and
it's like wow, hold on, this is enabling us to do really cool things that goes straight to the C-suite, right? When they say hey we're gonna do a campaign oh in a week, no problem, you know, so yeah, I, I would expect everybody's organisation to kinda, you know, notice good, you know, folks that are actually, you know, doing the right thing so and, and it seems like they have um.
So let's maybe look ahead to next year at this time when we're back here in Vegas for Accelerate. It's 110 degrees outside. What do you hope to be up here talking about next year? I don't know. I think there's a lot of work to do now, but, uh, you know, next year, I, I would like to see,
you know. I'd like to see kind of, see how the technology goes, right? Like, see, see what matures over time. Uh, sure, the virtualization is gonna get better, but, you know, I wanna see really quick, you know. Movement of workloads like containerized like just a more kind of operationally locked down
type we're gonna blast a lot of stuff out to this cloud provider and literally instantly move it to another one. So I, I wanna see like some more work go as far as like Kubernetti's like, you know, as a project just getting better at what it does, uh, getting a bit more, you know, nimble as far as, you know, moving, you know, workloads across the country around the world.
So right now it's, it's not static. I mean, it moves again with a good storage provider, you know, Port works and, and Pure, you know, it's fast stuff, right? So we can get 90% of the way there now, but you know, AI is a big thing, right? So uh I wanna see more, you know, essentially our vendors and software providers really
embrace the con containerization part, so just get rid of VMs. Like we don't need them anymore. Like, you know, we will for a little bit, but that's essentially where it's at, so, uh, some of you may be familiar with our pure storages proclamation that discs will be dead in 5 years.
Well, here hopefully VMs will be dead at Sirius XM, uh, within a couple of years and hopefully next year at this time we'll be up here talking about how we have, uh, essentially removed or eliminated VMs in favour of containers at at Sirius XM. Yeah, that'd be cool. Yeah, question.
What's your strategy for uh upskilling your VMware admins to become you know Kubernetes? Let me just repeat it. The question is, what's your strategy for upskilling your VMware admins to become Kuberneti's admins? It's still a VM, right? So there's training and there's, you know, knowledge transfer and,
you know, you would expect somebody to to turn around in a month and just change everything they know from, you know, one product to another, right? But there's a lot of similarities and what I've seen people do is like they, they're catching on pretty quick. They're like, oh, that's this feature.
So, you know, the names and the buttons change a little bit, but for the most part, the strategy is still the same, right? I think we get a little bit more, you know, out of our hardware. Uh, on Kubernetes versus something like VMware, but you know.
They, they need to wanna change themselves like you can't really ram it down, you know, their, their throat, right? They, they have to be excited about it. They need to learn about it. It's a lot of self teaching, but you know, for the most part you support them as much as possible, you know, bring vendors in to teach them, you know, different things and knowledge transfer
as much as you can yourself and some people just won't wanna do it, so it's like. Well, what are you needing any sort of resistance because I can imagine like a VMware admin who's been doing it for 3 decades and is supposed to retire because I don't. Yeah, yeah, we got a little bit of that in our broadcast group,
right? But, but they have a very important job and just get the content out there, so they're nervous and I, and I wholly, you know, understand like what where they're coming from. You don't wanna knock stuff off the air, right? So, uh, you know, they get comfortable with it, but over time they start poking around and you
know they'll get it after a while if they don't want to then. You know, they can manage the other platform until it disappears, so. And I think we're seeing a similar experience in in other enterprises as well. I mean, quite honestly that sort of skills and knowledge gap previously before the VMware disruption was big enough that people were not
gonna go tackle it, but now they've got a real economic incentive to to try to do something there um there are a lot of resources from folks like Red Hat if that's, you know, the kind of the the platform that you're planning to move to. Um, both educational, you know, certification, uh, as well as professional services offerings, they have a migration tool kit as well to actually migrate the VM workloads,
um, into the Red Hat OpenShift virtualization. Um, and then the other thing that we're hearing is that this starts to create a little bit of a, uh, modern environment that people want to come and work in, right? You, you sort of are able to attract that cutting edge cloud native, uh, modern talent when you're able to offer them an environment
that's more containers in in Kuberneti space, um, that varies right by different organisations for sure, but I will definitely say that the disruption has caused many. organisations to say yes it's a big gap but we're gonna find a way to overcome it through a number of the things that I talked about before. The economics are are sort of so strong, you know that it's really helping.
This is one of the things that's coming from the top down like no we're not gonna pay that, you know, we're just that this is insane, right? How did the the product increase in value over no, it didn't. It's the same thing. So why would we pay 10X? So C-suite's like, yeah, no, we're not doing it.
You know, you have, it's a mandate rather than a recommendation. So. We have time for one more question. I'm, I'm just kind of curious. I wanna rewind a little bit. You talked about prior to the broadcom announcement. platforms as well that that's costs, yeah,
like rev, yeah, sure, it was cheaper. Not at all. Yeah, I, I would say we started doing that like 5 years ago, you know, way before Bcom even entered the conversation and uh. Yeah, still, you know, get some value where you can, right? So I mean it's cheaper, sure, not fully, you know, as vetted out functionally,
you know, as far as an operator goes to be able to control it, but you know, again, you know, move some development stuff over there, you know, save, save, save a little bit here and there. And and it was actually a huge thing for us because it was like. When, when Red Hat said, we're getting rid of Rev, and you're going to OpenShift,
I was like, this is where I wanna go. Now I have an excuse. Oh, they're gonna sunset Rev. We, we gotta move. So it was definitely a loaded kind of thing, right? Because again, I, I just want the containers. I don't want the VMs,
so that gets us one step closer, so. Well, we're gonna draw this session to a close. I want to thank you all for spending the last 45 minutes with us. I want to thank Nate Mason from Sirius XM. How about a a round of applause for Nate and their modern virtualization journey.
We Also Recommend...
Personalize for Me