 Hello everyone. It's great to be here at KubeCon EU. My name is Vebhav Kamra and I'm the CTO here at Castin by Veeam. At Castin, we focus on data protection, so backup and recovery for Kubernetes applications. This is my sixth year here at KubeCon. It's really an amazing event. I love coming here because it's a great way to learn what the community has been doing as well as share experiences with with everybody out there in the same journey. So in that spirit, what I like to do today is talk a little bit about our experiences with our Kubernetes platform development as well as what we've observed working with a large number of customers as they've scaled out the number of users, the number of clusters, the number of applications they have and really the challenges they ran into. This is very much a here and now problem. If we look at the most recent CNCF survey, 80% of us who took the survey indicated that we are running Kubernetes in production. And what we're also seeing is that the journey from pilot to production, it's going much faster just because of the maturity of the tooling that's out there as well as all the learnings and education that's available to us. Now, security and business continuity are two key challenges that I wanted to focus on. How do we keep, make sure our environments are secure, both from internal and external threats from breaches. If vulnerabilities show up, how are we going to address them soon and fast? How are we going to detect those things? How are we going to deal with ransomware? So these are some of the key, what we call day two challenges, things that show up after you've gone through the initial stage of platform development. A lot of issues that we observe actually come down to misconfigured access. That is our users or tenants have too much access to the platform. There's this desire to enable self-service in these environments to allow our users to be able to do what they need to do without having to involve operations or file a ticket for everything. But what that can lead to is misconfigured policies. Privilege escalation is also very common. It's not just direct access to API resources that you have to worry about sometimes, but what access to other operators do users have in this environment? Do these operators that often have privileged access to the cluster, do they expose APIs that allow you to control who can access it or who cannot? Because if not, then users, there's going to be privilege escalation just because a user has access to one of the operator APIs that lets them do something that they shouldn't be allowed to do. The other issue that we see is the ability to react to vulnerabilities. So if there was a vulnerability discovered, if there's a zero-day exploit that needs to be, a patch needs to be rolled out for, can we do this fast enough without business, with minimum business impact? Because not being able to do that just leaves us vulnerable as well. And then lastly, the third one I wanted to talk about is ransomware. It's a big topic right now. It has a large financial and business impact. So do we have processes in place to detect ransomware as well as the ability to recover if ransomware hits and, for example, encrypts all our data? We want to be in the position that you're not held ransom to such an exploit. So what are the recommendations and learnings over here? So well, first, plan early on. We often talk about shift left in these environments. And what that means is planning and implementing processes that are required for production as early as possible in the lifecycle of the platform, in dev and test environments. You really want that to happen early. The second one is automation. Automation is key for everything we do in the cloud-native world. And it's no different over here. So having automation, for example, that builds over tooling like OPA, which is open policy agent for RBAC role creation, RBAC policy creation. That's something that we've seen our customers do and really helps with that misconfigured access issue that I had talked about. Having automation to roll out new clusters to migrate applications over from your old clusters that need to be patched over to these new environments, that's also key. You do not want to be in the situation or we don't want to be in the situation where we're stuck with environments that are vulnerable, that need to be patched. And we can't do that because there's a large business impact. So that's another recommendation over here. And then lastly, around business continuity, have a plan in place. Not just to deal with things like user error or disaster, but to deal with things like ransomware, having immutable encrypted backups, version backups available, so that you can deal with such issues is key over here. The basic principles here are not very different from our operational experiences in other environments. It comes down to planning, detection, mitigation, planning for this early on and automating all of the things that we just talked about. Thank you everyone for listening. I hope you have a great coup d'oeuvre. So thank you.