ec2.md

TL;DR

Wrappers for some AWS cli commands for EC2 Most often used to reboot a misbehaving worker node in AWS.

Use

filters

Some commands accept optional filters for selecting specific instances.

• --owner-id <aws account id>
• --private-ip <instance private ip>
• --instance-id <id of instance>
• <instance private ip> - same as --private-ip by default

The gen3 ec2 filter $filters command translates the above flags into the --filter arguments suitable to pass to aws ec2 describe-instances (see gen3 ec2 describe below).

asg-describe "default"|"jupyter"

Retrieve the json description of the "default" or "jupyter" node pool. You can distinguish between cluster nodes in the kubernetes cluster with the role label

Ex:

$ kubectl get nodes -l role=default
$ kubectl get nodes -l role=jupyter

$ gen3 ec2 asg-describe default

asg-set-capacity "default"|"jupyter" $number|+/-$increment

Set the desired capacity for the "default" or "jupyter" autoscaling group. If the number argument is preceded by "+" or "-", then increment or decrement the current desired capacity without violation the group max or min.

Ex:

$ gen3 ec2 asg-set-capacity jupyter +3

describe

Retrieve all info about instances

gen3 ec2 describe $filters

Options:

• filters: optional filters as described above

instance-id

gen3 ec2 instance-id $filters

Options

• filters: filters as described above

public-ip

gen3 ec2 public-ip $filters

Options

• filters: optional filters as described above

reboot

gen3 ec2 reboot $filter

stop

Stop an instance

gen3 ec2 stop $filter

snapshot

Snapshot the root drive of an ebs disk

gen3 ec2 snapshot $filters

terminate

Terminates an EC2 instance

gen3 ec2 terminate [-y] [--dry-run | -d] $filters

Options:

• filters
• -y: if present, it will terminate without requesting confirmation from the user
• -d | --dry-run: will do a dry run of termiantion (ie won't actually delete it)