Why use virtual computing?
Virtual computing allows users to access powerful and expensive applications from inside and outside of the library's physical location. Rather than waiting for the library to open to be able to use SAS or STATA, you can use those applications from the comfort of your own home.
What options do I have?
At UNC, there currently are two options for virtual computing: vcl.unc.edu and Virtuallab.unc.edu. Each option has pros and cons. Vcl.unc.edu is available anywhere, handles large datasets, allows access of local files, but is time limited. Virtuallab.unc.edu is available anywhere for unlimited time, allows access of local files, but does not handle large datasets well. Select the best option to suit your needs.
One option is virtuallab.unc.edu. This service runs a virtual application on the local computer, be it a computer at the library or your personal computer from the comfort of your own home. Once you've logged in to virtuallab with your Onyen and password, you may have to install the Citrix plugin, which works for both PC and Mac. Now navigate to the application that you want to use. Most of the statistical applications are in the aptly named folder Statistical Applications. Or you could alternatively just search for the application name using the search box in the top left side of the screen.
When you have the application loaded, there are a few things to keep in mind:
- Finding/loading files can take some work. When go to open files, remember the desktop in the window isn't the local desktop. You need to first click on the Computer button and from there you can navigate to where ever your files are located. Local drives are oddly labeled, but they are present. The C drive in this example is labeled Local Disk (C: on WI_nqwB-9qwsHUcHcR09).
You can also access thumb drives that are plugged in to the local computer and they will be labeled similarly to the C drive.
- As a follow up, when you're saving your work, be vigilant about where you are saving it. Do not hastily save to the desktop. Be sure to save it to the local drive or to a flash drive.
- In your code, to access local files you need to prefix the location of the dataset/etc with \client\. For example, if you want to access a dataset that is located at D:\ est.dta on the local computer then when using virtuallab.unc.edu you need to write \client\D:\ est.dta in your code so the virtual application knows where to pull the dataset from.
- The virtual application may get a bit laggy if you're doing something resource intensive. If you encounter lag then you are going to want to use the other service, vcl.unc.edu.
Your other option is vcl.unc.edu. This service connects you to another computer which has the application you want to use locally installed. This service too can be access anywhere so long as you have a valid Onyen and an internet connection. Once you've logged on using the Shibboleth (UNC-Chapel Hill) option with your Onyen and password, you need to create a new reservation by clicking on the New Reservation link.
Now you need to find the particular computer that has the application(s) that you need to use. Some of the descriptions are prefixed with brackets which makes scrolling through the list a bit of a chore, but the computers are labeled with the software packages that they have installed, however obliquely.
Then designate how long you want the reservation for. When your reservation runs out, that's it. So be sure to save your work before your reservation runs out. Wait until your reservation becomes available, click connect and then download the RDP file. After opening the RDP file to connect to remote desktop into the computer, you will have to enter a different password than your Onyen and that will be displayed at the same page where you get the RDP file.
You can load and save datasets and files from the local computer, but it's slightly different than with virtuallab.unc.edu. When you go to open/save the file, after clicking on the Computer icon, you will see the local drives listed. In this example, they are labeled C on HSL097 and so on. See image to the right.
Since you are running the program on the computer that you are remotely connected to, the processing power and/or memory of your local computer isn't the limiting factor so this ought to eliminate or mitigate the lag that you may experience when performing work on larger datasets.