Error in `enforceBundleLimits()` - package ignoring rsconnect.max.bundle.size option

[Lucas-Maciel]

Hi,

I'm having problems trying to deploy an app of 5Gb. The package is ignoring my rsconnect.max.bundle.size option and using the default.

> options(rsconnect.max.bundle.size=5000 * 1024^2)
> getOption("rsconnect.max.bundle.size")
[1] 5242880000

Error in `enforceBundleLimits()`:
! `appDir`
 (/Shiny/pancancer_datasets/T-cells/shinyApp)
  is too large to be deployed.
✖ The maximum size is 3145728000 bytes.
✖ This directory is at least 3460204092 bytes.
ℹ Remove some files or adjust the rsconnect.max.bundle.size option.
Backtrace:
    ▆
 1. └─rsconnect::deployApp(...)
 2.   └─rsconnect::listDeploymentFiles(appDir, appFiles, appFileManifest)
 3.     └─rsconnect:::explodeFiles(appDir, appFiles, "appFileManifest")
 4.       └─rsconnect:::recursiveBundleFiles(dir, contents = files, ignoreFiles = FALSE)
 5.         └─rsconnect:::enforceBundleLimits(rootDir, totalFiles, totalSize)
 6.           └─cli::cli_abort(...)
 7.             └─rlang::abort(...)
Execution halted

I tried different package versions, but I'm using v 1.0.1 on a MacOS now. Anyone has faced this before?

Thank you

[hadley]

How are you deploying the app? Are you running deployApp() yourself or using the buttons in the IDE?

[Lucas-Maciel]

Hi, I was using the buttons in Rstudio. With deployApp() it worked. Thanks

[aronatkins]

CC @kevinushey - one more example where the IDE deploy session lacks R options.

[kevinushey]

.rsconnect_profile is still the escape hatch we expect people to use in this situation though, I think? Or should we be making this accessible in an alternate way?

[aronatkins]

Ah! Thanks for the .rsconnect_profile reminder. :)