What are the best practices for storing and cataloging admin work?

Question

System administration

Over many years of practice, each system administrator has accumulated tons of developments - successful configs, direct procedures, time-saving workarounds, and so on. Many of them go to the community and add to our common knowledge base. Some, due to their confidentiality or other reasons, remain only in local copies and often end up in all sorts of subdirectories with romantic names like "customer_N/etc" or, say, "to_sort_out_asap"
And as a result, you again spend an extra 30 minutes (in best case) in Google to solve a long-solved problem and engage in sports cycling.
What is the best approach to storing, tagging, searching through your work? Who uses what? Is there anything better than plain text files, find, grep and vim?

Answer 1

Weird why no one has suggested a wiki yet.
You write articles for yourself, link to each other, configs can be inserted into the page in pieces, or you can attach the full config to the attachment. Access, again, can be separated to whom what, and to whom nothing, plus the history of changes is kept who changed what.
I use dokuwiki for my work, mediawiki at work. And either is good. Plus configs of working servers can be put in git/svn.

Answer 2

a local tracker, where I set myself tasks and a local svn, where all scripts, configs, etc. are stored. I have been carrying this turnip with me for years ... eleven :)

Answer 3

I started a blog, where I describe difficult situations or a rake that I had to deal with. When the question pops up again, I climb and read how I struggled the previous time :-)

Answer 4

Tree of folders with files.
Each folder contains a text file containing the necessary tags separated by a dash in the name.
Inside the file - you can make any mark / description.
Total:
1. Folder - has a normal name of what's inside.
2. File name - associative tags and description inside.
Then: do a search for a tag by file names using the "mask": [*desired tag*.tags.txt]
Result: destination folder with full path - subject + title, file - these are tags.
I've been using it for a long time, no problems yet.

Answer 5

ansible/chef/puppet+git

Answer 6

Public on the blog, private on the Wiki system, portfolio on Instagramm, "wisdom" on Facebook, scripts on Github. I have a link to Jira under this Confluence. It is very convenient to write technical specifications for something in Confluence and immediately set tasks in Jira from there.