London FTW!

I've just created a new Linode in the London DC and cloned my original Linode to it. Before you can clone a Linode it needs to be shut down, so I shut it down and initiated the clone.

Once the clone is initiated you can power the original linode back up, but that made me wonder why I had to shut it down in the first place. I can imagine a few things, but why could I fire it back up again while the cloning was in progress?

Oh well, inpatient as I am, I fired the original up expecting that since this is possible a snapshot of the disks is being copied over to London, but I don't know for sure... And I like being sure...

So: should I learn to be more patient or will this Just Work™

While a VG is snapshotted, all writes are amplified into a read and a write, and then another write. This, frankly, sucks, so we try to avoid snapshots when possible. Cloning does not use LVM snapshots, so you really should keep the source Linode powered off during the cloning process.

Hope that helps,
-Chris

That helps

The cloning is queued on the new Linode so I've issued a boot job for the new Linode. But I cannot queue the boot of the original Linode since there are no jobs in the queue. Since a powered down Linode is more or less a requirement maybe you could make it so that a full Linode clone takes up a spot in the queue or add the option to power the VM back on after the clone completes in the clone form?

But in my case since I am moving my Linode to London, creating a migration ticket will have the same effect...

This is the best reason to use Beta Backup. You can create a snapshot and restore it to another (temporary) node in the same data center, then clone it.