Am 13.07.2017 um 10:23 schrieb Arturo Borrero Gonzalez: > On Wed, 7 Jun 2017 13:35:13 +0200 Moritz Muehlenhoff <j...@debian.org> wrote: > >> A couple of possible solutions, but these are all rather something for >> upstream development: >> - sysctl.conf files could gain an additional parameter which specifies >> the kernel module creating the sysctl. systemd-sysctl could then >> trigger the loading of that module and wait until the specified sysctl >> is present (I think that's the cleanest solution)
I think the proper solution is to find out why the /sys entries are not available after the module has been loaded and fix that in the conntrack module. systemd-sysctl.service already has a After=systemd-modules-load.service ordering. >> - All sysctl settings should be entirely idempotent. systemd-sysctl could >> be re-run in multi-user.target (fairly ugly) That would still be racy, as the module that has been loaded has still not created its sysfs entries. >> - Or maybe split off a separate systemd-sysctl-stage2.service, which reads >> a separate sysctl config (which can then be used for driver which are >> loaded >> late) Would still be racy, see above. > > Hi, > > I'm fine with any of them. > It would be great if systemd maintainers could comment on the solutions. > To me this smells like a kernel problem, specifically the conntrack module and should be addressed there. But this would need input from a kernel maintainer. -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
signature.asc
Description: OpenPGP digital signature
