Salting things up in the sysadmin's world
summary
SaltStack is Python based and uses YAML and Jinja2 and the talk shows concrete howtos: master/minion architecture, sls state files, top files and highstate runs that deploy packages and templated configs. It builds a custom Python grain to detect multicore systems which syncs to minions and drives matching, and it demonstrates using pillar for centralized variables and templates. Notes cover node groups, matching strategies, Foreman integration and practical limits such as pillar encryption tradeoffs and log collection.
AI-written summary · about
chapters (42)
- 0:00Introduction
- 0:35Who am I
- 1:13What is configuration management
- 2:20Quantity
- 3:43Lazy people
- 4:56Salt
- 5:48Terminology
- 7:11Environment
- 8:33Top File
- 10:03State
- 13:05HiState
- 14:28Matching
- 14:55Note Groups
- 15:57No Groups
- 16:18Grains
- 18:12Pillar
- 19:15Pillar Antioch
- 20:05Pillar Data
- 21:10Questions
- 21:26Twoparter
- 22:34Anode configuration changes
- 23:08Master debugging
- 23:44Salt is evolving
- 24:27Beginner questions
- 25:04How would I do that
- 25:34Why Salt
- 26:10Monitoring
- 26:45Secure connections
- 27:14Language
- 27:29Puppet
- 27:49Pinal
- 28:26Pilar encryption
- 28:51File encryption
- 29:34Threat model
- 29:55Encryption
- 31:01Salt vs Puppet
- 32:31Managing Dependencies
- 33:09Include State Files
- 34:15Split by Files
- 35:13what happens if the grains raise an exception
- 35:50when is the code actually executed
- 36:27how do you manage logging
related talks
52mScalable Systems Management with Salt
40mSalt: How to be Truly Lazy
58mIntroduction to
29mLive Demo
19mWhy did I create Salt?
51mBuilding Web Scale Using Open Source on the Internet Assembly Line- 19mWhy did I create Salt?
39mBeyond Open Source Configuration Management: Rapidly Building Any Cloud at Scale