Stopping and resuming datomic cloud instances


#1

I’ve just tried to setup datomic cloud in “Solo” setup for playing with it.
I’m wondering if it’s possible to stop and later resume running instances while preserving the state of the system.
I tried to stop EC 2 instances (small and nano) manually, but it looked like the main instance was “terminated”.
Then I started bastion instance again but it seems that two completely new instances have been created. However, I cannot connect to the bastion anymore (possibly because of wrong keys?).

At this point, I don’t need to run instances all the time - only now and then for quick experiments. I’m therefore curious to hear what are your suggestions how to reduce costs in such scenario.


#2

As you noted, if you terminate the compute instance of your Datomic system the Autoscaling Group will create a new one to replace it. This behavior is by design, to maintain the system in case of an instance failure.

You can always delete your compute stack, as documented here, without losing any data.

However, if you want to keep your stack but stop all instances, you can also edit the Autoscaling group rules for the group named <your-system-name>-TxAutoScalingGroup:

  • Find the ASG in the Autoscaling Console,
  • highlight it and click “Edit” in the lower dialog box
  • change “Min” “Max” and “Desired” to 0
  • click “Save”

The ASG will now terminate your running instance and will not replace it until you change the scaling parameters back to 1.


#3

Great, exactly what I was looking for. Thanks!