bash-scripts

log-archiver

files: log-archiver.sh, log-archiver.json

Dependancies:

• You're running this script from a user that has permissions to all servers & files.
• If planning to run this script from cron, This user owns the crontab you wish to use.
• The connection to external servers are already tested and secured with a gpg key.
• If using AWS, you have already configured a connection and using a key for authentication.
• You've already got jq installed. If not, get it from apt or a similar repository for your version of linux.
• A JSON has already been created with the information this script will use.

Description:

In order to get a good data lifecycle going, we need to first harvest the data, prepare it for archival, and get it to the cloud. That is what this script enables. It does this by creating an ssh connection to a list of servers, copying over the target files (in this example, logs), and merges them into a compressed tarball (tar.gz). It will then add a log of its activity to the tarball, and if desired, send a copy to an S3 bucket. I made this script to accept all input parameters from a JSON file. That way, the configuration parameters may be programically generated in the future. This script seemed especially useful for starting a backup of my home envrionment, which is why I created it.

What I learned:

1. How to use jq. This thing is super handy!
2. How to get some basic error handling and file validation in my scripts.

git-interface.sh

Description:

This is a simple text interactive interface for some of the basic functions of Git. I made this using VS Code, tied to my copy of WSL2 which is running Ubuntu with Git installed. This seemed like a good idea to help refresh myself on Git and build out my bash portfollio.

What I learned:

1. I learned how to build a menu. Nifty!
2. Prompt Statements and how to use the right one for the job.
3. Creating the repository on the github side requires using the API as well as an Oauth style key. It may be easier to go to the website for one-offs.
4. No matter what directory manipulation you do in a script, once the script ends, you're back to the directory you started in.
5. QUOTES MATTER IN GIT API!! "{"name":"$repo_name"}" is correct. '{"name":"$repo_name"}' is not.

filemaker.sh

Description:

This short little script takes a few arguements and generates a file with the desired name. It is scalable, using Arguement1 you can specify any number of files you want made. The first file will be named whatever Arguement2 is. Successive files will be named Whatever Arguement2 is, but will also have a number apended to the end. So if you wanted three files named test, you would get back "test, test2, test3". I made this because it's actually a good example of a more complex for loop.

What I learned:

1. Getting a for loop to be both an index and a variable is tricky and c-like. https://tldp.org/LDP/abs/html/dblparens.html
2. Spaces in the if statement are super important, especially those spaces around the square brackets.
3. Semicolon at the end of the if statement appears to not be needed. However if you had it, you could move 'then' up to that line.