This is merely a historical archive of years 2008-2021, before the migration to mailman3.
A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/gerrit-log@lists.osmocom.org/.
Neels Hofmeyr gerrit-no-reply at lists.osmocom.orgPatch Set 1: Code-Review-1 (1 comment) https://gerrit.osmocom.org/#/c/2443/1/contrib/jenkins-run.sh File contrib/jenkins-run.sh: Line 39: ssh "$osmo_gsm_tester_host" "ln -s $trial_dir $osmo_gsm_tester_dir/trial-latest" There's an issue of "polluting" the namespace of trials. This trial dir serves as a name and trial.py assumes it to be a unique entity. Note that trial.py is capable of finding a trial that has not yet run, for a possible daemon that automatically discovers new trials that showed up (see http://git.osmocom.org/osmo-gsm-tester/tree/src/osmo_gsm_tester/trial.py#n39 ). 'trial-latest' would then reference a trial a second time. trial.py should be able to tell that the trial-latest has already run, but there should be a test for that. So, what if I launch a manual test run on 'latest', and then the jenkins swaps out the symlink? AFAICT the osmo-gsm-tester would continue to use the new path and choke up. Ok, we can add this patch *if* you also make sure that a symlink gets resolved to its actual path before a trial starts. So Trial.__init__() should employ os.path.foo to resolve the link, http://git.osmocom.org/osmo-gsm-tester/tree/src/osmo_gsm_tester/trial.py#n52 and *if* you make sure instead of the link name, the resolved name is used everywhere. The "latest" link is more of a convenience for us to be able to invoke the latest manually, right? The user could also use 'ls -1t | head -n 1'? -- To view, visit https://gerrit.osmocom.org/2443 To unsubscribe, visit https://gerrit.osmocom.org/settings Gerrit-MessageType: comment Gerrit-Change-Id: I26ddf55110738bd1944ccbfe72e8410ff9811392 Gerrit-PatchSet: 1 Gerrit-Project: osmo-gsm-tester Gerrit-Branch: master Gerrit-Owner: Pau Espin Pedrol <pespin at sysmocom.de> Gerrit-Reviewer: Neels Hofmeyr <nhofmeyr at sysmocom.de> Gerrit-Reviewer: neels <nhofmeyr at sysmocom.de> Gerrit-HasComments: Yes