RStudio Connect 1.7.8 – Put a pin in it!

September 22, 2019
By

[This article was first published on RStudio Blog, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

This release adds new workflows for data scientists and improved production
settings for administrators. For data scientists, it used to be hard to use the
same data or R objects in different content, and even harder to update those
resources regularly. This release enables you to pin objects in Connect to solve
these challenges. For administrators, we’ve reduced the most common sources of
publishing failures and significantly improved error handling. All together,
version 1.7.8 makes it even easier for data scientist teams to share and
leverage their work with the enterprise.

Updates for Users

Experimental Support for Pins

Pins Support in RStudio Connect

The pins R package provides a way for R
users to easily share resources using RStudio Connect. Your resources may be
text files (CSV, JSON, etc.), R objects (.Rds, .Rda, etc.), or any other
type of files you want to share. Sharing these files can be useful in many
situations, for example:

  1. Multiple pieces of content require the same data. Rather than copying
    that data, each piece of content references a single source of truth hosted
    on RStudio Connect.

  2. Content depends on processed datasets or model objects that need to be regularly updated.
    Rather than redeploying the content each time the information changes, use a
    pinned resource and update only the dataset or model. The update can be automated using a
    scheduled R Markdown document. Other deployed content will read the newest data on
    each run.

  3. You need to share resources that aren’t structured for traditional tools
    like databases. For example, models saved as R objects aren’t easy to store
    in a database. Rather than using email or file systems to share these R objects,
    use RStudio Connect to host these resources as pins.

Refer to the RStudio Connect user
guide
or the pins
website
for more
information.

Example APIs for Model Serving

API Monitoring in RStudio Connect

RStudio Connect makes it easy for data scientists to share models written in R
with other teams as RESTFul APIs. As part of the 1.7.8 release, we’ve expanded
our documentation to help teams approach model management with the following
examples:

  1. The end-to-end use of RStudio Connect to train, deploy, monitor, and A/B test a model.

  2. Add additional logging to API requests in order to track latency, performance, input parameters, and route popularity.

Updates for Administrators

Deployment Error Logging

Better Error Logs

We’ve overhauled the RStudio Connect deployment process for R code. RStudio
Connect now captures errors and surfaces specific error codes and
recommendations. All publishing methods see these improvements including
push-button publishing, Git-backed deployment, or custom workflows using the
Connect Server API.

A glossary of the error codes and recommendations is available here.

R Package Repositories

By default, RStudio Connect attempts to install R packages from the package
repositories that were used in the development environment. In some cases, you
may wish to specify different behavior and tell RStudio Connect where it should
look for R packages:

  • Your developers install packages from a public CRAN mirror, but your production server must use an internal CRAN mirror.
  • You use an isolated network for your production server, so RStudio Connect can not access the package repository used on the development network.
  • You want to use RStudio Package Manager’s package binaries in production.

The Admin Guide describes how to configure the package repositories RStudio Connect should use for R.

Usage Scorecard and Feedback

Usage Scorecard in Connect

Like you, we’re committed to expanding the influence of data science in the
enterprise. A new usage scorecard on the Admin Dashboard’s Metrics page helps
you understand how your team uses RStudio Connect and what additional
capabilities may be available. We have made it easy for you to share this
scorecard with RStudio, providing feedback that will help us further improve
RStudio Connect.

Deprecations and Breaking Changes

  • SECURITY: New Timeout Defaults – The Authentication.Lifetime and Authentication.Inactivity settings have new default values that are in line with industry best practices. These settings control how frequently a user must refresh their log in to RStudio Connect. The new default values are shorter ensuring users are authenticated more frequently.

  • BREAKING CHANGE: External Package Check – On start up, RStudio Connect now checks to ensure every R packages listed as external in the configuration file is available in every version of R on the system. Any missing packages will cause RStudio Connect to fail to start. This check prevents unexpected deployment and runtime failures for content. You can opt out of this check by setting Packages.ExternalCheckIsFatal to false.

  • BREAKING CHANGE: R Markdown Rendering Errors – R errors that occur during R Markdown rendering now stop the deployment process with an error. Previously, R errors would result in the error message being rendered as part of the document contents. The previous behavior can be restored by setting chunk options for the R code chunk in the Rmd file, e.g. {r error=TRUE warning=TRUE}.

  • PLATFORM SUPPORT: Trusty – RStudio Connect 1.7.8 is the last release that will support Ubuntu 14.04 Trusty. Please refer to the RStudio Platform Support policy for more information.

  • PLATFORM SUPPORT: REHL 8 – RStudio Connect 1.7.8 is the first release to support Red Hat Enterprise Linux 8. Refer to the RStudio Connect Admin Guide for more information.

  • BREAKING CHANGE: These previously deprecated settings have been removed; see the release notes for more details: LoadBalancing.EnforceMinRsconnectVersion, Applications.ExplicitPublishing, Authorization.UsersListingMinRole, and Password.UserInfoEditableBy.

  • DEPRECATIONS: The following settings have been deprecated and will be removed in the next release; see the release notes for more details: Applications.DisabledProtocols, OAuth2.AllowedDomains, and OAuth2.AllowedEmails.

Refer to the full release notes for
more information on all of the changes and bug fixes in this release.

Upgrade Planning

Please take special note of the breaking changes above, especially the new check
for external packages. Aside from the deprecations and breaking changes above,
there are no other special considerations and upgrading should require less than
five minutes. If you are upgrading from an earlier version, be sure to consult
the release notes for the intermediate releases, as well.

Get Started with RStudio Connect

If you haven’t yet had a chance to download and try RStudio
Connect
, we encourage you to do so.
RStudio Connect is the best way to share all the work that you do in R (Shiny
apps, R Markdown documents, plots, dashboards, Plumber APIs, etc.) with
collaborators, colleagues, or customers.

You can find more details or download a 45-day evaluation of the product at
https://www.rstudio.com/products/connect/.
Additional resources can be found below.

To leave a comment for the author, please follow the link and comment on their blog: RStudio Blog.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.

Search R-bloggers

Sponsors

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)