August Reading: how to run software?

I’ve been reading about how to run software. Here are some of the more curious approaches that I’ve examined:

• the debathena way
• Russ Allbery’s Debian + puppet way
• the Nix way
• the Wikimedia Operations way
• the Akamai way:
  • “The Akamai Network”
  • ACMS
  • Query
• the Google way:
  • scheduling: “Omega: flexible, scalable schedulers for large compute clusters”
  • configuration: “GCL Viewer: A study in improving the understanding of GCL programs”, 2008 Masters Thesis by Ibrahim Bokharouss
  • deployment: “Maintaining Consistency in a Massively Parallel Environment”, a USENIX UCMS’13 presentation by Dinah McNutt
  • isolation:
    • “CPI^2: CPU performance isolation for shared compute clusters”
    • “NicPic: Scalable and Accurate End-Host Rate Limiting”
  • build: the Google Engineering Tools 2011 “Build in the Cloud” blog series:
    • “Accessing Source Code”
    • “How the Build System works”
    • “Distributing Build Steps”
    • “Distributing Build Outputs”

Sidenote: Google probably aren’t using much of the regular Linux userland, are they?