Dear users of the RStudio Server service at LRZ,
we have to inform you that the service in its current form will have to be shut down by November 30th, 2021.
We are working on a new service associated with the LRZ AI Systems that will become a suitable replacement for most use cases. If you're interested in being added to the waiting list of early testers for this service, please prepare a short summary of your use case and workload, including a specific description of system resource requirements, i.e. how many CPU cores are needed, what is the expected RAM requirement for your workload, do you currently/plan to utilize GPU accelerators for your work? Submit your request to the LRZ Servicedesk (with the "Description" form field "RStudio Replacement Waiting List Request"). We will then contact you with further instructions once test access can be made available.
Please also contact us via LRZ Servicedesk if you have any additional questions or concerns.
What is RStudio Server?
RStudio Server provides a web-based integrated development environment (IDE) for R (as well as Python and other programming languages). It can be accessed using a browser on basically any device with internet access. It offers a convenient interface for developing and running R (and other) scripts on LRZ systems. In addition to an R console and editor, it provides, amongst other features, a file browser for analyzing results and plots or transferring files between the LRZ Linux Cluster and your local device.
In order to use the RStudio Server at LRZ please visit www.rstudio.lrz.de
When asked to authenticate, please use your LRZ Linux Cluster account to log in.
RStudio Server at LRZ is a service associated with the LRZ Linux Cluster. If you already have an account for the Linux Cluster, you can use the corresponding user name and password to access the RStudio Server. Otherwise, you will have to request access to the Linux Cluster first (see Access and Login to the Linux Cluster)
Linux Cluster Integration
RStudio Server provides a file browser which can (in the default setup) be found as 'Files' tab in the bottom right pane of the web interface. It can be used to access the data in your Data Science Storage (DSS)-backed Linux Cluster home directory ($HOME).
Any other DSS based storage containers - including the new DSS-based Linux Cluster project containers - can also be accessed using RStudio Server. To do so, the data curator of the DSS container has to apply the LRZ RStudio Server Container Config to the DSS container in question. The general procedure is described here (section 12).
You can also use the built-in Terminal to submit jobs to the Linux Cluster's batch queues via the Slurm Workload Manager (see Job Processing on the Linux-Cluster).
R Package Management
You can install your own R packages in the usual way using the install.packages()-function or the 'Packages' tab in the bottom right pane of the web interface. RStudio will ask you whether to install these packages into a personal library which it will, per default, create in your Linux Cluster home directory ($HOME). Beware that this is a different package library than on the Linux Cluster login and compute nodes. Make sure to install needed packages in both environments if you plan to submit jobs to the Linux Cluster compute nodes.
History of Notable Changes to the Service
2021-01-19 RStudio Server Upgrade
RStudio Server was updated to the latest release of version 1.3 (for a feature overview see https://blog.rstudio.com/2020/05/27/rstudio-1-3-release/), however not all features will be available initially but may be activated at a later time.
2020-07-20 R Version Refresh
Available R versions were updated and the default version changed:
- R version 3.6 was bumped to version 3.6.3 and set as default for new sessions
- R version 4.0(.2) was made available
- R versions 3.4(.4), 3.5(.2 and .3) were retained for backwards compatibility. These may be removed in the future. If possible, please use newer versions.
2020-01-29 Service Maintenance and Upgrade
On January 29th, 2020 this service underwent maintenance work. During this maintenance, the virtual infrastructure of the RStudio Server was moved to a new back-end. The following changes need to be taken into account by users of the service:
- Available system resources were upgraded considerably. Please adhere to the new usage limit of 20 CPU cores in total per individual user at any given time. These can be utilized in a single or multiple R sessions. You may run a maximum of 5 concurrent R sessions.
- The operating system of the RStudio Server nodes was updated to Debian 10. This requires you to re-install your individually added R packages. Please contact us if you encounter problems (e.g. missing system dependencies) when adding your custom packages.
- RStudio Server was updated to version 1.2 (for an overview see https://blog.rstudio.com/2019/04/30/rstudio-1-2-release/), however not all features will be available initially but may be added at a later time.
- The default version of R was updated to 3.5.3 and should work with most user packages (while older and newer versions are available for you to select if needed).
- Read-only access to the legacy NAS-based file-systems is not possible anymore. Home directories based on Data Science Storage (DSS) continue to be accessible automatically. In order to access additional DSS project containers, please see the relevant information documented above. Submission of jobs to the Linux Cluster compute nodes will continue to work as expected.