 Welcome to our session on providing disaster recovery for Rosa clusters. I'm Patrick Ubele technical marketing engineer with redhead and I'm We're Rico technical business development manager at net up It's nice to meet you all and I give you a quick introduction to astra control and Lewis will present the Interesting bit a quick demo on how to protect Rosa clusters with astra control Astra control is a multi cloud service fully managed by net up it provides application data protection for your rest data full application kubernetes applications so you can have disaster recovery and application portability and Possibility to migrate your applications to other clusters other clouds It's supports most of the kubernetes distributions Including open-shift and Rosa of course And how does it work from a high level so the astra control service? Will discover all the namespaces and applications in your managed Open-shift clusters in that case if you trigger a backup of a defined application Astra control can make can make use of execution hooks to freeze your application. So stopping the IO for a short moment Then it will back up the application configuration and its metadata to astra controls internal database in the next step it will initiate snapshots of the Persistent volumes that host your your persistent data. So these can be backed by either net up storage or the hyper hyperscalers persistent storage So the snapshot operation is quick once the snapshot is done Astra controller will unfreeze the application so it will resume normal operation and then in the background it will start Copying the data to external object storage Once you have both the snapshot and the external object storage in place You can either restore your application from snapshot if you do it with the same cluster or Should you have a disaster like your cluster has been wiped out for whatever reason or you want to migrate the application to another cluster or Even a different cloud you can restore the application from the external backup hosted on object storage and with that There is no Patrick so Either to do a demo and we have only six minutes for the demo. So I will try to speak very fast. So It's a recorded demo. Okay, and we are this is our demo environment. So Here we are what we are doing is protecting and providing disaster recovery for Rosa cluster But this can be done also for open sieve anywhere. Okay, so let's start the demo and see if this is working That's super Yes, so the demo is using one Rosa cluster called Rosa main in the island region and we also use Dear Rosa cluster in the north Virginia region So we have a postgres sequel name a space to to be protected a stateful application This is using some pbc and astrakontrol service is a manage service in the public cloud So it connects to these two Rosa clusters use the cube config and And it's just using Kubernetes APIs. Okay, so then we need and Sorry, I'm going to fast here So we need also an object store to store our work back up. So all the backups are stored externally So even if the entire cluster goes down, we can still recover things and we are doing these backups as Patrick explain using CSI Snapshot so astrakontrol servers will do a snapshot do a backup from the snapshot and restore and restore it in the in the new Rosa DR cluster so we can so now how this Rosa cluster is in the in the in the AWS console Okay, and you will see the instances belonging to the Rosa cluster the Rosa main cluster in the island region and also the Instances belonging to the Rosa DR cluster in the north Virginia region. Okay We will also show the the open sieve console. So here it is you have the the Rosa main Instances, I hope the the size is well for you to see and if we switch to north Virginia You will see the the instances of the Rosa DR cluster running in north Virginia and in that AWS region we have also the red hat open sieve Console so is Rosa we are running version 4.12. This is the Rosa main cluster We have also the web console for open sieve for the Rosa DR And if we go back to the Rosa main cluster, we can see that we have this deployment and this spot running Okay, it's a deployment config using postgres sequel and we have also the persistent volume claim for for that deployment config Okay, all of this is captured in the astra control service web console. Okay, in this case, we are using EBS, okay, and this is the the astra control service Web console. Okay, so here this is the dashboard when you enter your credentials You can see in the account that I'm using my own account And I can provide credentials to enter the main impress scalars And I'm also using cube configs to access the Rosa main and Rosa DR We have this bucket section So we need obvious stress to do the backups to store the backups I have a bunch of buckets to be used as the destination of our backups And we have also the two clusters the Rosa main cluster that is already Registered for data protection in the astra control service and the Rosa DR cluster So as you can see they are running 125 that is 4.12 if we go to the Clusters the Rosa main you can see the bunch of the name and spaces and Including the postgres one and here we have some information about the worker nodes and the control plane nodes in this Rosa cluster Okay, if we go to the applications, we are already protecting our postgres sequel Stateful application. It's a deployment config if we go inside the postgres sequel You can see that we have snapshots and backups for this application And we are doing daily backups because we have a configuration a configure a protection policy We are doing backups at 9 a.m. Every morning and we have doing we are doing snapshots Every hour at the 30th minute of the hour. Okay, so everything is a schedule So we have a protection policy and it's taking care of protecting the postgres sequel application for us And all these snapshots are CSI snapshots We can see the PVC that we are protecting through with CSI snapshots But above all all the Kubernetes resources all the metadata that we are protecting every time we do a snapshot every time We do a backup. Okay, so we are saving off all these custom resource or the config map the secrets The deployment config definition the service account So almost everything that is related to the postgres sequel to this name a space We are saving it outside every time we do a backup. So just to do Disaster recovery is as easy as going to the actions button and just click in we have we can do a snapshot backup clone and just click in restore and There is this wizard you can restore to the original name a space You can restore to new name spaces when you select restore to new name spaces You can select the destination cluster of the That you want where you want to restore to and I have selected the rosa d r cluster We can change the name of the destination name a space for the postgres sequel database So I'm using postgres sequel d r. Okay. I just click next and Then you can see you can select from which Snapshot or backup we want to do the restore. Okay, so in this case I'm using the latest backup available to do my disaster recovery Okay, and then you just have to check the summary and click restore. Okay, so Going to the terminal window behind the scenes you can see that here in the top window we have the rosa main, you know cluster. Okay, and We have this deployment config in the postgres sequel project name a space, okay with the pbc and everything and And in the bottom window, we have our disaster recovery cluster. These are the open sift The projects and name spaces that we have so if if we search again for the name spaces in the disaster recovery You can check that There is a new name a space called postgres sequel d r that it was just created 25 seconds ago And if we go inside and see what is happening inside this This project this name a space we can see that the postgres sequel is getting Restore, okay From the from the s3 bucket. Okay, so you can see they are the pbc that is already created and a restore Job that is happening at the same time if we wait a Couple of minutes we will see that the the restore is completely done. Okay, so it's not important to see How much time is taking or not the the the disaster recovery process? The the important thing is how easy is to do a disaster recovery using this tool No, so it's next next next so even my cat can do a disaster recovery of an open sift application With astra control service. Okay, so and and that's all okay So this is the demo if sorry if we have no time if you have any question We can we are available at lunch and and and thank you very much for listening I'm sorry for a life demo come to our booth. We can show it