-
Notifications
You must be signed in to change notification settings - Fork 13
Update Workflow's with newer versions of Postgres, Ubuntu, and Checkout action #440
Conversation
Test failure is odd - 255 error is an ssh issue. What is ginger exactly? |
Actually I don't know if I'm right on that ssh part, need to do a bit more googling on the error |
So this bash is a little murky to me. In # resolve path to this file (following symlinks) and load libs
BASEDIR=$( dirname $( ${CMD_READLINK} -f "${BASH_SOURCE[0]}" ) )
for f in ${BASEDIR}/bin/cli.sh ${BASEDIR}/bin/* ; do source $f; done which is calling this https://github.com/NYCPlanning/db-pluto/blob/main/pluto_build/bin/cli.sh |
Ginger is the ftp we get that data from, the process seems to be running totally fine on main. I think the error might be coming from the latest version of ubuntu - let me switch that back to 20.04 and see if it runs |
@fvankrieken worked when I reverted back to 20.04 |
found this resource related to the issue: actions/runner-images#2791 I am going to see if that fixes the issue with the action |
the current pts run seems to have gotten past the earlier issue, it is no no longer throwing the mysterious 255 Error Code |
Very strange. If things are running fine now, I would say good to merge= |
I messed up and ran the PTS workflow with the 20.04 version of ubuntu and it worked so I actually can't get the workflow to work with a newer version of ubuntu, I think it makes sense to just leave it at 20.04 and monitor these workflows when we rebuild PLUTO. Does that sound good to yall? |
Addresses issue #439.
Minor PR to address maintenance task to update frequently used tools in our github workflows to the "latest and greatest". Pretty simple stuff, the only thing to note is that I added an option to increase the shared memory capacity to
--shm-size=2gb
as the PLUTO build is quite large and the default 64gb of shared memory was not sufficient to build PLUTO. The time to build PLUTO was reduced by about 10 minutes with updated versions of PostgreSQL (performance improvements between versions).Testing the workflows:
I tested a few of the workflows, and the runs are here: