Well, a bit of a segway. Thanks to your efforts our tests are far more stable and this has allowed me to increase testing's workload (the reason runs went from 3.5 to 2.5 hours is that I upped the # parallel boots). However, it's also exposed what I suspect is yet another problem with 9p. On very rare occasions - we're talking once every 5 or so runs - a test fails to boot. For instance:
https://testing.libreswan.org/v3.28-1473-g1b0f13544a-master/nflog-01-global/OUTPUT/ <cagney> [ 6.345150] FS-Cache: Netfs '9p' registered for caching <cagney> [FAILED] Failed to mount /testing.\r <cagney> See 'systemctl status testing.mount' for details.\r <cagney> [DEPEND] Dependency failed for Local File Systems.\r <cagney> [DEPEND] Dependency failed for Mark\u2026 need to relabel after reboot.\r ... <cagney> You are in emergency mode. After logging in, type "journalctl -xb" to view <cagney> system logs, "systemctl reboot" to reboot, "systemctl default" or "exit" <cagney> to boot into default mode. I guess I should file a bug. However, I'm not sure where; and I'm not sure how helpful the above is for details :-) On Sat, 25 Jan 2020 at 04:36, Paul Wouters <[email protected]> wrote: > > https://virtio-fs.gitlab.io/ > > Will be available soon. Might be good candidate to replace 9p > > Sent from my iPhone > _______________________________________________ > Swan-dev mailing list > [email protected] > https://lists.libreswan.org/mailman/listinfo/swan-dev _______________________________________________ Swan-dev mailing list [email protected] https://lists.libreswan.org/mailman/listinfo/swan-dev
