Even though the database backup files are created and sent to the S3 bucket daily, eventually the bucket will have too many files. So I need to delete the older copies of the database files

ps: This shouldn't be done by date but by number of items in the buckets.

pps: consider versioning in S3

Install SSL certificate on the EC2 instance
This might be handy: https://sectigostore.com/page/steps-to-install-an-ssl-certificate-on-aws-ec2-instance/

Currently using t2.micro

Consider using,

• t4g.nano (free till 31st March 2021 but uses ARM processor)
• t2.nano (almost half the price)

After recent tests, the server instance rarely uses over 300mb of the 1GB t2.micro ram
Switching to a 500mb ram shouldn't affect a wiki site but would cut cost to almost half

In case we lose everything, read the following and check if we have everything we need to restore it,
https://www.mediawiki.org/wiki/Manual:Restoring_a_wiki_from_backup

If we don't have certain files/dependencies for a full restore, create it as an issue.

If the instance goes down, what is the fastest way to create a new instance with all the user data already installed and wiki database connected?

Find ways to automate as much of this as possible. That involves, not only installing,

• Apache
• MariaDB
• PHP
• imagemagick
• mediawiki (delete downloaded file later to free up server space)

but also

• moving the database onto the server and plugging it into mediawiki

Would need to create a list of deliverables and test it out before creating related issues on github.

This should also include things you would do differently this time around