User_data change forces new resource. about terraform-aws-ec2-instance

[naomichi-y]

I have the same problem...
can't add user_data to lifecycle.ignore_changes?

[jeffrey-irwin]

Issue open with the provider for this: (hashicorp/terraform-provider-aws#23)

[tbugfinder]

Hi @antonbabenko ,

would you support adding a lifecycle block as input?

Thanks

[antonbabenko]

@tbugfinder I just answered in another issue - #166 (comment)

[krystan]

ok so I should mention there is a workflow hack to avoid another resource, you can change your actual resource user data and then do a state refresh, the ec2 instance meta data will be written back to your state with the new hash avoiding the forcing of a new instance you cannot do a state push with a new sha hash if you do a refresh as the hash will be calculated from the existing resource (terragrunt does this) and the statefile repopulated.

[bryantbiggs]

hi all - the hash change will destroy and re-create anew as it is currently written by the AWS provider. as previously discussed, lifecycle hooks currently do not support variables and as such we would be making strong assumptions for the entire user base of the module. therefore, at this time there isn't a solution for this other than creating your own fork of the module and injecting your own lifecycle hook to ignore changes to user_data

[krystan]

Seems we need to wait for variable support, we forked and use where appropriate

[bryantbiggs]

closing for now as we cannot conditionally support this lifecycle setting

[github-actions]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues. If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.