osmo-gsm-tester.git/contrib/jenkins-build-srslte.sh cannot be used
because there's a need to split the fetch step from the build step.
That's because we want to fetch in the host (in order to use host's user
ssh key for potential private repos) while still building inside the
docker container, where we have all the dependencies (and where the
binaries are going to be run afterwards by osmo-gsm-tester).
During the changes above it was also noticed that /tmp/trial on the host
got its UID+GID changed (because we switch user inside the container to
downscale privileges). Fix it by chown'ing the directory back to the
host user's UID+GID.
Change-Id: Ia0db8fba840392a0b7883cfc15e45c2e80407a16
After this patch, one can run osmo-gsm-tester with specific parameters
(debug level, suite and tests to run, etc.) by using
OSMO_GSM_TESTER_OPTS variable:
OSMO_GSM_TESTER_OPTS="-s 4g -t ping -l error" ./jenkins.sh
Change-Id: I2e8ff51a3152125148c6a117a987e8c11a576e1e
* Fix wrong KI.
* Add an extra IP address to slave so that ENB and EPC can be run on
different IP addresses and their GTP ports don't collide.
* Add required rf parameters to resources.conf
Change-Id: I0198bc99f75e9567ee717aa79e434ec28705fac9
This set up allows running a typical osmo-gsm-tester set up with a main
unit (ogt-master) attached to one (or more) slaves where resources are
located and processes are run.
As an example, current config allows running an iperf3 test over an srsLTE
stack using the ZeroMQ backend.
Change-Id: Iac7c0b613048ce332642d3957c55cc7c53b56d42