Snapshot automation for droplets and volumes in DigitalOcean Cloud platform

Digtalocean Automate Snapshot for Droplet and Volumes

Reference https://pypi.org/project/goutte/

Issue - For Python 3.7 on Centos 6

Reference - https://benad.me/blog/2018/07/17/python-3.7-on-centos-6/

Genrate Api KEY for digital Ocean via

https://www.easydeploy.cloud/blog/automate-droplet-snapshot-digitalocean/?unapproved=39&moderation-hash=6864fe13b1a4810007aa7067bbdca7db#comment-39

only first 2 Step to be followed

Finally on Live Server For example , at mail.blsinternational.net

goutte /usr/local/etc/goutte.toml 36c095a48c01428e14b1a38c3ddce5cb57d65df42cb1d727e59a22735b1a6236 --only snapshot

And file /usr/local/etc/goutte.toml

retention = 2     # Number of backups to keep per droplet/volume
[droplets]
names = [          # Array of droplets you want to snapshot
 'mail.blsinternational.net',
]
[volumes]
names = [          # Array of volumes you want to snapshot
 'bls-zimbra-store-01',
]

And Crontab Entry is

0 17 * * * goutte /usr/local/etc/goutte.toml 36c095a48c01428e14b1a38c3ddce5cb57d65df42cb1d727e59a22735b1a6236 --only snapshot > /dev/null

Deletion of Snapshots automatically keeping last 2 old Snapshots

Please install doctl via https://github.com/digitalocean/doctl

Create the following script

[root@mail ~]# cat /root/scripts/delete-old-snapshot.sh

#!/bin/bash
doctl compute snapshot list | grep goutte | head -n2 | awk '{ print $1 }' | \
while read i
do
 echo "$i"
 doctl compute snapshot delete -f $i
done 

As all the snapshot of BLS international starts with goutte , we have used the logic of grepping this key word

In Crontab keep the time as 1 hour after the creation script has executed

0 18 * * * /root/scripts/delete-old-snapshot.sh >/dev/null 2>&1 

Keep checking if it working fine periodically . Snapshots are costly so you should not have snapshots for which customer is not paying .