Publish application to RStudio Connect failed

Our company recently purchased RStudio Connect and got it installed. I tried to publish a simple API , which has dependency on ROracle, to our RStudio Connect. But the publishing failed with the following error:
[Connect] RequiredFile: 'OCI libraries'
[Connect] Package: 'ROracle'
[Connect] PackageVersion: '1.3-1'
[Connect] Found in the following log entry:
[Connect] 2021/10/07 18:51:06.387482176 configure: error: OCI libraries not found
[Connect] Build error: An R package cannot be installed because a required system-level dependency is missing. (r-missing-system-library)

Our RStudio Connect has R4.1.1 installed, which is the same version I am using in RStudio IDE. And I already installed ROracle on RStudio Connect server, and tested it. It has been working fine.

This problem is a blocker of my project. I need help to get it fixed. Thank you very much for any advices.

Hey @albert ! Welcome to RStudio Community!!

It's worth noting a couple of things:

  • The "OCI libraries" are an Oracle thing. As far as I understand, no connection to Oracle can succeed without these libraries (1) installed on the system and (2) configured properly with whatever library you are using
  • ROracle is woefully un-maintained - last updated in 2016
  • We have seen several of our customers be more successful (though it can still be tricky to configure the OCI libraries properly) with the odbc package and an ODBC connection to Oracle: GitHub - r-dbi/odbc: Connect to ODBC databases (using the DBI interface)

In particular, since you are using RStudio Connect, RStudio provides professional drivers which include an Oracle driver and should help you get set up connecting to your oracle database!

A specific guide on installation is here:

https://docs.rstudio.com/pro-drivers/installation/

Once you have the drivers installed, you should be able to configure an odbc connection (and even get it working outside of R with a DSN, isql, or otherwise). Then it should be easy to make use of within R! There are lots of ways to make this possible, many of which are discussed here:

https://db.rstudio.com/best-practices/portable-code/