You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Invoking /init via bash seems to eliminate the error about /run. After doing that, another error occurs later on when Rstudio server tries to increase the number of files it can open. I believe this is because my user does not have the associated capability for changing that. But I noticed the limit on my system was already much higher than what Rstudio wanted, so I simply commented out that line of code in the Rstudio sources, rebuilt the deb package, and then rebuilt the rocker/rstudio image with this modified Rstudio.
I'd be interested to hear if this approach may be improved on somehow (or is way off the mark), and also any insight into why invoking /init via bash seems to be important.
The text was updated successfully, but these errors were encountered:
works on my system. The --ulimit setting eliminates the need to patch and rebuild RStudio as I did before. My guess is that updates to podman have eliminated the need to add bash -c /init to the command line. I think users interested in using podman would do best to use a recent version. I have had success with podman version 1.6.2 on Ubuntu 19.10 and podman version 1.7.0 on a recently updated Arch Linux system.
I seem to have gotten rocker/rstudio to work with podman on Arch Linux by making a few small changes. Here is my command line
Invoking
/init
via bash seems to eliminate the error about/run
. After doing that, another error occurs later on when Rstudio server tries to increase the number of files it can open. I believe this is because my user does not have the associated capability for changing that. But I noticed the limit on my system was already much higher than what Rstudio wanted, so I simply commented out that line of code in the Rstudio sources, rebuilt the deb package, and then rebuilt the rocker/rstudio image with this modified Rstudio.I'd be interested to hear if this approach may be improved on somehow (or is way off the mark), and also any insight into why invoking
/init
via bash seems to be important.The text was updated successfully, but these errors were encountered: