Learning Library

← Back to Library

Accelerating Innovation with Stateful Containers

Key Points

  • Organizations prioritize “speed to market” by building the simplest, fastest solutions with the highest chance of success.
  • A forward‑thinking firm chose a hybrid‑cloud strategy that deploys all critical apps and data via containers, balancing both stateless and stateful workloads.
  • For stateful containers, five key requirements are emphasized, starting with persistent storage that can handle structured, file‑based, or unstructured data across blob, file, or object stores.
  • The architecture must preserve application configuration and ensure consistent behavior when moving workloads between on‑premises and cloud environments, enabling rapid scaling, mobility, and temporary container lifecycles.

Full Transcript

# Accelerating Innovation with Stateful Containers **Source:** [https://www.youtube.com/watch?v=MiCK6_jZigo](https://www.youtube.com/watch?v=MiCK6_jZigo) **Duration:** 00:12:26 ## Summary - Organizations prioritize “speed to market” by building the simplest, fastest solutions with the highest chance of success. - A forward‑thinking firm chose a hybrid‑cloud strategy that deploys all critical apps and data via containers, balancing both stateless and stateful workloads. - For stateful containers, five key requirements are emphasized, starting with persistent storage that can handle structured, file‑based, or unstructured data across blob, file, or object stores. - The architecture must preserve application configuration and ensure consistent behavior when moving workloads between on‑premises and cloud environments, enabling rapid scaling, mobility, and temporary container lifecycles. ## Sections - [00:00:00](https://www.youtube.com/watch?v=MiCK6_jZigo&t=0s) **Accelerating Innovation with Hybrid Cloud Containers** - The speaker outlines how a forward‑thinking company leverages hybrid‑cloud architecture and containerization—balancing stateful and stateless workloads—to maximize speed‑to‑market, mobility, and rapid, disposable compute resources. ## Full Transcript
0:00organizations today that are looking to 0:02be competitive fast-moving Etc know one 0:06really important thing and that is speed 0:08to Market what the market is looking for 0:11is for organizations to build the 0:13fastest simplest thing that they can 0:16that's going to have the highest 0:17probability of success let's take a look 0:19at a Forward Thinking company one who's 0:21looking to rethink the way that they do 0:23Innovation one who's really thought long 0:26and hard about not just the kind of 0:28architecture they want to put in place 0:30hybrid Cloud but also the way that they 0:32want to deploy applications and data 0:35this particular organization has decided 0:38that they really want to take full 0:39advantage of speed and have as such 0:42decided that they are going to deploy 0:45all of their critical applications and 0:48data sets via 0:54containers they know that there are some 0:56applications that they have that are 0:57going to be stateful in nature we're 0:59going to focus on on those but there'll 1:01also be different application Stacks 1:02that they'll run that may be stateless 1:04as well so as they think about the kind 1:06of hybrid Cloud architecture that 1:07they're looking to deploy they have to 1:09keep in mind that the architecture needs 1:11to support the statefulness of the data 1:14set the configuration of the container 1:16and the overall architecture of the 1:17hybrid Cloud platform it's running on as 1:20well as enable for Speed and Mobility 1:23those stess containers that need to just 1:25pop up to do some work and then they can 1:27disappear after it's over whenever we're 1:30talking about stateful containers one of 1:32the things that we have to think about 1:34and there are actually five of them are 1:36just the very first and the simplest 1:38service that any stateful container 1:39requires which is the ability to be 1:45persistent so as the container itself 1:48lives and breathes it amasses data as 1:51it's processing information and taking 1:53inputs from the users that support it 1:55this particular company may do online 1:57retailing for example and they know that 1:59there are certain applications that will 2:01benefit from having the persistence by 2:04which the information that's stored will 2:07be available the next time the 2:09application is started they also know 2:11that the data will be placed on the most 2:14efficient storage platform within the 2:16cluster itself that it's capable of 2:18working with data sets whether they are 2:20stored in blog storage file storage or 2:24in object storage in other words is the 2:27data structured in nature that it can be 2:30inter intersected with a row and a 2:31column um is it file based or is it 2:35something that's completely semi or 2:37unstructured in nature and has no 2:40addressable row and column intersection 2:43all of these kinds of considerations the 2:45container and its persistence needs to 2:48be able to 2:49maintain not just that but also the 2:52configuration of the application itself 2:54so that if the application needs to move 2:56from the on print location to the cloud 2:59then that consistency that experience 3:01that you're looking to maintain at a 3:04consistent way transfers with the 3:06application it could also be in a hybrid 3:09Cloud manner that the application maybe 3:11it runs in the cloud and it's accessing 3:14parts of its data set on Prem some parts 3:17in a completely different Cloud 3:18environment and maybe some parts in the 3:20cloud that the application is running on 3:23this idea of persistence allows the 3:25application to know exactly where it was 3:28at the last point the user interacted 3:30with it and it has the underlying 3:32services that address the storage 3:34requirements whether that's snapshotting 3:37backups provisioning expanding 3:40Contracting encrypting all of the things 3:43that a persistent storage container 3:45would need to be able to provide for the 3:48application that's running in a stateful 3:50way now speaking of persistence the 3:53other thing that this particular company 3:54needs to keep in mind is 3:58resiliency 4:02I mentioned a couple of times that the 4:03application may run on Prim it may run 4:06in the cloud it may run and across 4:08multiple clouds a combination thereof 4:11the ability to provide High availability 4:14of the application itself is a component 4:16of resiliency should the application 4:18fail this approach or these Services 4:21would allow the application to quickly 4:23be restarted to minimize the amount of 4:25downtime and to certainly mitigate any 4:28data loss we often refer to that as data 4:31breaches so the notion of resiliency 4:33isn't about just the application in the 4:35container it's also about the prevention 4:38of breaches of that information so for 4:40example maybe I typed um you know an RM 4:43star. star at the root level of my 4:47particular container host well be great 4:51if I could get that back without having 4:53to reload the application and start all 4:55over from scratch and go and try to find 4:57that data set itself so this notion of 4:59being being able to be resilient isn't 5:01just about the application it's not just 5:03about the data it's also the resiliency 5:06services that are necessary to prevent 5:08data loss from accidental activities and 5:11to certainly make it so that the 5:13application can quickly be restarted so 5:16that there's no downtime Associated by 5:19the enduser who's taking full advantage 5:21of the application itself now after 5:24resiliency it stands to reason that we 5:26also need to be considerate of security 5:32open up any web browser go to any news 5:34site and I'm sure you'll find no 5:36shortage of stories that are associated 5:38with a service that was disrupted by 5:40some threat actor or some nefarious 5:43state who's trying to do things that 5:45they should not be doing and so the idea 5:48of security here is how do we prevent 5:50the threat actors from getting into the 5:52system to begin with this is where we 5:54take full advantage of services that are 5:57baked into the platform that the 5:59container is running on to protect the 6:01data in the application in the event 6:03that there is some kind of an attack 6:05there's a rapid response that you can 6:07put into place even before the response 6:10A system that allows you to detect 6:12activity that might be occurring on the 6:14network that could be indicative of a 6:16threat attack itself to correlate that 6:19Network related activity with what's 6:21happening to the data especially the 6:23relationship between the host and the 6:26actual data source itself for example if 6:29a host has normally been sending data in 6:32the clear and all of a sudden it starts 6:34encrypting the information well that 6:36could be one of two things it could be 6:38that we've just changed our security 6:40policy and we've turned on host side 6:42encryption it could also mean that the 6:44host has been impacted by some kind of a 6:46threat that seeks to um not just maybe 6:48encrypt the data but maybe even to 6:50exfiltrate the information 6:52itself now correlation becomes critical 6:55here so the ability to ask the other 6:57parts of the platform for which this 6:59container is running on top of if 7:01they're seeing indicative signs of 7:04different kinds of threat activities 7:05would allow us to remove the false 7:07positives so the security components 7:09here that are protecting the data 7:11snapshotting backup and Recovery 7:13traditional backup and Recovery tools 7:15are complemented now by the activities 7:18that are monitoring and trying to detect 7:20as well as services that allow us to 7:23repair any kind of damaged information 7:25automate the response if you will of an 7:28attack so that the recovery activities 7:31follow the defined runbook that you've 7:33set forward and they recover in an 7:36autonomous fashion or manual if that's 7:38the route that you want to go all of the 7:40services that you see listed right here 7:42surrounding this container that I've got 7:44drawn in the middle need to be able to 7:47follow the container and the underlying 7:49data wherever it resides whether it's on 7:51Prem in the cloud across clouds or some 7:54combination of all of 7:58that 8:00the key to Mobility is consistency being 8:03able to package all of the things up 8:04that I just said that are defined as 8:06policies that complement the container 8:08and the underlying data so that wherever 8:10this resides it has the exact same 8:12experience that's delivered in the 8:14destination that you're trying to drive 8:15it to Mobility is not just a one-way 8:19location or a one-way uh form of 8:22Transport if you will bidirectional 8:23Mobility is what's critical here 8:26especially when you're talking about 8:27applications that are designed designed 8:29in vehicles that really support a hybrid 8:32Cloud 8:33architecture that same experience that 8:35we might have inside of a public cloud 8:38or some kind of cloud hosted location 8:41means that the application that I might 8:44have up 8:46here in the cloud will have all of the 8:49same capabilities that we've listed 8:51already the persistence the actual 8:54resiliency the security and now the 8:57mobility that b directional Mobility 9:00that we're talking about means that as 9:02the application moves from an on Prim 9:04environment all the way up to the cloud 9:06it should be able to move in the exact 9:09same manner back to the host on 9:13Prem now that's a really critical 9:15concept because the last thing you want 9:17to do is create and develop up in the 9:19cloud environment and then be stuck 9:21there because you can't get the 9:22application out there happens to be one 9:24or two services that are running up 9:26there that aren't running on Prem so 9:29when you think about the kind of 9:30platform that you want to build this 9:33particular container architecture on top 9:35of it needs to be able to run as 9:39application stack running on Hardware uh 9:41or as a software stack that's running up 9:43inside of the cloud 9:45itself so we've talked a lot about what 9:47we've done with the container itself we 9:49also want to make sure that the 9:49container has access to the actual 9:52storage um data sets that it actually 9:55needs to use to be 9:58successful 10:01so the last thing that any Forward 10:04Thinking organization is going to be 10:05thinking about is how do I make sure 10:07that the applications that I'm running 10:09in my container platform have access to 10:12all of the data sets that are available 10:14to it that it should have access to 10:17provided to it in a way for which it's 10:19been enriched and it's been cataloged 10:21meaning I know where the data is I know 10:23what the data contains and I know the 10:26degrees of relevance and value it has to 10:28the organiz ation and this particular 10:30application I can use that cataloging to 10:33affect policies how often it gets backed 10:35up who has access through different 10:37kinds of control measures authentication 10:40and authorization considerations could 10:42be baked into that I can set governance 10:45and maybe regulatory policies Based on 10:47data contained within the catalog such 10:49as how good this data is uh and can be 10:52used for as well as who has access to it 10:55can it even go to the cloud or does it 10:57have to stay on Prem all of these things 11:00really make the overall containerization 11:02experience of an application not only 11:05consistent but also more relevant to the 11:07organization it creates higher degrees 11:09of accuracy within the application 11:11because it has access to all of the data 11:14it's supposed to regardless of where the 11:16data is stored in the organization and 11:18in the format it may be stored in 11:20remember these containers and the 11:22applications that they support need to 11:24be be able to access data that could be 11:26stored on block based systems on file 11:29based systems and on object based 11:32systems as well that notion of creating 11:36the experience that's consistent on Prim 11:38means that you're going to make specific 11:40decisions in your containerized 11:42architecture that effectively deploy 11:45Cloud architectures on premise to where 11:48the storage the compute and the 11:49networking simply become services that 11:52are almost invisible just like they are 11:55in the cloud they're certainly there the 11:58container need them and we can use 12:00policies to adjust the kinds of 12:02performance requirements that that 12:04particular application demands and the 12:06container environment requires to 12:08support that application experience as 12:10we need to within the cluster that we've 12:13created thank you if you like this video 12:16and you want to see more like it please 12:18like And subscribe if you have any 12:21questions please drop them in the 12:22comments below