We talk to Candida Valois, field chief technology officer (CTO) at Scality, about the legacy applications and modern, cloud-native environments that are driving the need for object storage.
In this podcast, we look at organisations with requirements to keep large volumes of unstructured data both online and available, while also protecting it from security threats. At the same time, modern application requirements, built via the principles of DevOps and as cloud-native, are emerging. All of this, says Valois, is driving object storage.
Valois talks about the suitability of object storage to storing large amounts of unstructured data and successfully protecting it.
Antony Adshead: What’s driving the growth of object storage right now and how much of that is to do with the rise of microservices and cloud-native?
Candida Valois: I see two groups creating a lot of the growth. Some of them with exact requirements, but other requirements are different.
The first one is your classical legacy. I believe that every industry we deal with is now very digital.
We’re talking about financial institutions, we’re talking about medical, we’re talking about genomics. You cannot find an exception to that rule. The data is so valuable now that people are paying millions of dollars to protect it. And some people need to keep the data online longer now than before.
It started with media and entertainment because they tend to repurpose and monetise it, but now you can see that across all the different industries. Across banking, healthcare, genomics, government and others.
You can’t find an industry that can’t value their data and the long-term value of the data. As a result, they are looking for a solution to store it, to keep the integrity of the data, to protect it; all the different threats. We’re talking about unplanned events, typical system errors and what’s happening in the past six years or so [such as] malware and ransomware.
So, you have the classical, typical industry. Now, we are seeing the modern applications also consuming storage.
Modern applications require a different type of storage. You can see a real shift in the past four years or so. Now we’re seeing a wave of new workloads that we can call cloud-native.
We’re seeing a significant emergence of big data applications, things like analytics. You can see that blending with artificial intelligence, machine learning and … the interesting thing about the deployment of those models for those applications is not only the datacentre, but also in the cloud and at the edge.
So, that is a big change. Those applications are looking for performance, they are looking for some sort of demand dynamic search engines.
And you can say it is even changing the footprint of the storage. Especially in the entry point, you don’t have to start with the high TB [terabytes]. For many of those use cases you can start with tens of TBs.
That’s what we’re talking about with microservices that are using Kubernetes, that is completely different from the legacy world. We’re used to the storage administrator, the storage expert with theirs hands on the storage but now with the microservices, the DevOps, the DevOps team, the user, they are the ones doing it. They want a completely different model and it is creating a lot of data as well.
Adshead: Where do the obstacles and opportunities for object storage lie?
Valois: It’s interesting. A few weeks ago, I saw a research report that said that 95% of businesses said the need to manage unstructured data is a problem for their business.
Think about that just for a second; 95% of businesses are saying that. And you also have to think about that 80% of the data of those businesses is unstructured data.
So, this is a big issue. I believe that that’s the main challenge right now. They have to protect it and they have to keep it for a longer period of time. And they want to do all that right. And if you drill a little deeper into the customer and what exactly they want, they want to do that with more, and they want to do more but with less.
So, to do that, what do they have to do? They need to improve their staff efficiency, they have to reduce their silos, but at the same time they need data to be protected, they have to keep it online and it has to be accessible.
And, furthermore for those new applications, those application owners they want to deploy something very quickly, very agile without the dependency on the storage administrator as an example.
So, that is the big challenge that I see. However, as a result, that provides an excellent opportunity for object storage.
Object storage provides that scalability. It provides protection, with data available from everywhere, any time that they need.
Also, I believe object storage is the right solution for those challenges because it supports both scenarios that I mentioned before. It can support those legacy use cases: for big data, backup, content management, media and entertainment.
And at the same time it can provide the support for cloud-native use cases like AI/ML, things that you have to analyse for those DevOps teams and DevOps users.
And furthermore, with the evolution of flash devices, you can even have scalable object storage utilising a full flash environment at a low cost.