Hi! I have a question about Linstor. I installed it as a storage solutions for k8s but its HA architecture is not clear to me. So, the base layer is DRBD and linstor is on the top of it. I have 3 linstor nodes (one combined and two satellites). If one of the two satellites goes down, the cluster will keep working, there is no question about it. However, what will happen if the controller (in my case combined) node goes down? How is the cluster going to operate without the controller?
Hello Samuel! If a LINSTOR controller fails, your data is still accessible! You are simply unable to schedule new volumes until your controller is back up and running. If you would like to cluster the controllers, you can take a look at our blog here: www.linbit.com/en/linstor-controller-pacemaker/. Soon, our Operator will take care of rescheduling failed controllers and replicating LINSTOR's internal database via ETCD. If you would like to learn more, we’re happy to schedule a consultation call: just send a message to sales@linbit.com including the country that you are located in and we can direct you to the correct team.
Wow, man! Please do more! You have a great way of explaining. Your redundant network diagram was very helpful!
Great video, man. How can someone thumbdowned this great explanation of HA?
Excellently explained!
Hi,
I was working on Always on 2016 server and one of my Node went into quarantine state.. can you help me with this ..
Hi!
I have a question about Linstor. I installed it as a storage solutions for k8s but its HA architecture is not clear to me. So, the base layer is DRBD and linstor is on the top of it. I have 3 linstor nodes (one combined and two satellites).
If one of the two satellites goes down, the cluster will keep working, there is no question about it. However, what will happen if the controller (in my case combined) node goes down? How is the cluster going to operate without the controller?
Hello Samuel! If a LINSTOR controller fails, your data is still accessible! You are simply unable to schedule new volumes until your controller is back up and running. If you would like to cluster the controllers, you can take a look at our blog here: www.linbit.com/en/linstor-controller-pacemaker/. Soon, our Operator will take care of rescheduling failed controllers and replicating LINSTOR's internal database via ETCD. If you would like to learn more, we’re happy to schedule a consultation call: just send a message to sales@linbit.com including the country that you are located in and we can direct you to the correct team.