cloud engineer
Instead of using AWS and JQ to get the snapshot names, you can do it with a single command using –query.
Here are the previous commands.
/usr/bin/aws lightsail get-instance-snapshots --region us-east-1 --profile default > $snaps cat $snaps | jq -r '.instanceSnapshots[] | .name' > $names |
Using query.
aws lightsail get-instance-snapshots \ --query 'instanceSnapshots[*].[name]' \ --region us-east-1 \ --profile default \ --output text > $names |
When trying to boot from a disk created from snapshot and you’re getting an “UEFI feature is not available for this project” error, you’ll need to create a boot image with the –guest-os-features option set to “UEFI_COMPATIBLE.”
gcloud compute disks create new-server-disk-2 \ --project=project-name \ --source-snapshot=server-disk-2 \ --zone=us-central1-b \ --guest-os-features="UEFI_COMPATIBLE" |
It’s about enabling guest operating system features on custom images. The guest operating system features let you configure networking, security, storage, and operating system options on custom images to be used on boot disks.
Here’s how to describe snapshots using query with tags.
aws ec2 describe-snapshots \ --owner-ids xxxxxxxxxxx \ --profile default \ --region us-east-2 \ --output text \ --query "Snapshots[*].[SnapshotId,Tags[?Key=='Name'].Value[] | [0]]" |
Result.
snap-xxxxxxxxxxxxxxxxx Server1 snap-xxxxxxxxxxxxxxxxx Server2 |
Here’s how to copy a disk from one project to another.
gcloud compute disks create lamp-server \ --source-snapshot https://www.googleapis.com/compute/v1/projects/project-id/global/snapshots/lamp-server \ --project next-project-id \ --zone us-central1-a |
Once all the disks are copied. You can create a snapshot of the boot.
gcloud compute disks snapshot lamp-server |
Create an instance from snapshot.
gcloud compute disks create "hostname-boot" \ --project "project-id" \ --zone "us-central1-a" \ --source-snapshot "snapshot-name" \ --type "pd-standard" \ --size "100" |
Here’s my script to delete AWS LightSail snapshots. It can delete daily or weekly backups and scheduled via crontab. If you want longer or shorter retention, adjust the expired value. It’s in seconds. 604800 is 1 week. 2592000 is 1 month.
#!/bin/bash current=$(date +%s) if [ $# -eq 0 ]; then exit 1 fi if [[ $1 = 'daily' ]]; then prefix='daily' expired=$(($current-604800)) elif [[ $1 = 'weekly' ]]; then prefix='weekly' expired=$(($current-2592000)) else exit 1 fi snaps='/root/snapshots/snapshots.json' names='/root/snapshots/names.txt' parse='/root/snapshots/parse.txt' logfile='/root/snapshots/snapshots.log' /usr/local/bin/aws lightsail get-instance-snapshots > $snaps cat $snaps | jq -r '.instanceSnapshots[] | .name' > $names cat $names | grep $prefix > $parse while read -r line; do snapshot=$(echo $line | cut -d_ -f3) snapshotname=$line if [ `expr $snapshot + 1 2> /dev/null` ] ; then if [ $snapshot -le $expired ]; then echo 'Deleted: '$snapshotname >> $logfile /usr/local/bin/aws lightsail delete-instance-snapshot \ --instance-snapshot-name $snapshotname else echo 'Nothing: '$snapshotname >> $logfile fi else echo $snapshot is not numeric > /dev/null fi done < $parse echo 'Current time: '$current >> $logfile echo 'Expired time: '$expired >> $logfile echo '-----------------------------------' >> $logfile |
Schedule deletes via crontab.
# run daily at 5am 0 5 * * * /bin/bash /root/snapshots/delete-snapshot.sh daily 2>&1 # run weekly every sunday at 6am 0 6 * * 0 /bin/bash /root/snapshots/delete-snapshot.sh weekly 2>&1 |