Skip to content

Instantly share code, notes, and snippets.

@BlackMaria

BlackMaria/TerraformStateCorruption.md

Last active May 18, 2016 14:12
Show Gist options
  • Save BlackMaria/1fb25962b050a7ccb4b9f2ff22563007 to your computer and use it in GitHub Desktop.
Save BlackMaria/1fb25962b050a7ccb4b9f2ff22563007 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
TerraformStateCorruption.md

The Issue

  1. When attempting to create a stack, if the connection gets cut (or the user cancels the command inflight), some requests to AWS may have succeeded after the command ceased writing to the log.

  2. Upon the next run of the command, terraform will fail because the "names" of the "resource" will have already been created.

  • This requires the manual removal of those "resources"

The fix? -- Suggestions

  1. Before sending a request to AWS, save what is about to be requested.
    • It is also suggested to give a temp name (a 128bit hash?) and rename after create succeeded
    1. The next time we run, if there are requests that were not complete, we can then check to see if it had succeeded (This is why I would suggest starting with a temp name)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment