48 Hours ago we had an issue with a bad transistor on our of our nodes, after the replacement of the defective part the Raid configuration of the node went totally bad rendering boot impossible. After multiple hours working on it I was able to bring back the Node to normal activity. Such incident is rare and should not occur.
In order to counter that we had a brand new node deployed in our Newest DC which is equipped with multiple benefits :
- 10 GBps connection speed for routers In/Out traffic.
- Ability to move the Vms between our different Nodes at 10GBps speed minimizing any downtime by doing live migrations of data.
- Ability to backup Virtual machines once a weak.
- Better backup once a week is provided free of charge with a full month retention of images.
- Ability to take snapshots of Virtual machines will make it possible to roll-back any actions done on the VM to the version prior-backup
- Vm migrations between nodes will provide ability to keep the provided IP address.
- Better version of Libvrtd for virtualization
- Better monitoring and fault tolerance Tanks to the Raid10 Disk setup offering higher IO and consistency with Samsung Enterprise SSD.
What will happen in the upcoming hours ?
We will begin the migration operation for your server. This means that we will have to shut down your server during the operation. VM migrations at the opposite of data migration means that all data will be safe and be Consistent.
What to expect: The only difference and change that will be applied is the IP change. Once your server back online, we will bring it back using VNC to change various networking settings in order to restore connectivity on it.
What you have to do?
Once the server moved you will just need to replace your old IP with the new one. This should not take long to propagate. We advise you to lower your A record TTL to make it shorter.
