How should I handle a Scheduled Migration for my Linode?
I received a ticket that informs me to anticipate Scheduled Maintenance on my Linode. Unfortunately, the time they planned to happen is terrible for my business. What can I do to avoid any downtime?
✓ Best Answer
When we schedule a migration for your Linode, it's related to improvements we are making to our infrastructure. Improvements include adding new features or updating hardware like host machines and networking gear. It's important to remember that we serve an international customer base, making it nearly impossible to ensure that a scheduled migration will occur outside everyone's business hours.
These migrations cannot be avoided or delayed. If you would like to avoid downtime that will occur during the Linode's scheduled time, here are available options:
You could add the Backup Service to create a backup of your Linode. From there, restore that backup to a new Linode. Once you've determined the restored backup is running as expected, you can transfer the IPv4 address between the two Linodes. It's important to note that IPv6 addresses cannot be transferred between Linodes. With Network Helper enabled, all that would be required is a reboot of the two machines to bring up networking. Meaning, the only downtime you'll face is the time for the reboot. You also avoid needing to update most DNS records pointing to your Linode.
- If you rely on the IPv6/AAAA records, you need to update the DNS record with the new IPv6 address.
- This does involve an additional charge. We invoice the Backup Service on an hourly basis. Once you're up and running, you can remove the service to avoid accruing additional expenses.
Clone your Linode
Similar to the Backup Service, you can create a copy of your Linode by cloning it to a new Linode. Where the Backup Service is a paid service, cloning is not. The only charges incurred are related to the additional service added to your account.
- It's highly recommended that you shut down your origin Linode to prevent data corruption on the cloned Linode.
- The downtime needed to clone a Linode is related to the data usage on the Linode. Larger Linodes and Linodes with large file counts will take longer to clone. There may be cases where the clone could be a similar or longer time than a migration.
You could schedule a time that works best for you and the users accessing your Linode to kick off the migration early. You can kick off the migration at any time by clicking on the Enter the migration queue now. link on the Linode's dashboard in the Cloud Manager. Your Linode will then power down and enter it into the migration queue. Doing this will allow you provides more control and visibility into the migration and can respond should anything go awry with the services starting on the Linode.
Setting your services up with a High Availability is a good idea beyond the scope of just avoiding downtime for a migration - especially if your services are mission critical. This can help in cases of emergency maintenance or a server unexpectedly going down. A service like our NodeBalancers is a great way to get started with that, though there are many options out there. Yes, there's time needed to consider the best options for your workflow and then time required with setting HA up, though maybe a scheduled migration, with enough of a heads up for the deadline, can be the motivator needed to get this set this up.