[BUG] PRD tries to access UAT - bucket & fails

[SanderDevisscher]

Describe the bug When deploying to the PRD we get an error (see screenshot) which raises suspicion that PRD server tries to access the UAT - bucket for data checks and fails in doing so.

To Reproduce Steps to reproduce the behavior:

1. Deploy docker in PRD environment
2. open app (@berthuygens can you confirm ?)
3. See error

Expected behavior app works in PRD

Screenshots

[mvarewyck]

@berthuygens Could it be that the environment is misspecified in the shinyproxy yaml file (I don't have access to it)?

For UAT it should have sth like R_CONFIG_ACTIVE = "uat" While for production this should be R_CONFIG_ACTIVE = "production"

See our config file See shinyproxy configuration container-env

[mvarewyck]

I had a look at the yaml files and they seem correct, so I have no clue why for production it looks in the wrong bucket

• Is it failing for both the public & private (wbe) app in production?
• Can you try again with my latest changes here? I updated the code to start an alternative app when it fails to start reportingGrofwild::runWildApp() which prints the configuration settings: the current configuration name, bucket and whether there's a datacheck.
• Could you send me the log files? Maybe I can see from there why it goes wrong.

[stenmigerode]

Hi @mvarewyck,

This problem was fixed on Monday. This ticket should have been closed

[mvarewyck]

@stenmigerode Can you please specify how this was fixed? - for future reference

[stenmigerode]

@mvarewyck There was a problem with the deploy on Friday, which was fixed later on. That is why it tried to access the UAT bucket.

Also the IAM Roles for accessing the PRD S3 bucket from the PRD Shiny Proxy server were not yet configured previously, so we added those policies in AWS.