Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On Tue, 2016-10-11 at 14:59 +0200, Boris Fiuczynski wrote: > > I tried upgrading from 2.2.0 to 2.3.0 a bunch of times, but > > I haven't been able to reproduce the failure you're > > reporting: libvirt-guests is never started automatically. > > > > Can you provide more information? What distribution are > > you using? Are you building from source, or using your > > distribution's packages? Are you sure libvirt-guests was > > not running even before upgrade? > > I used fc20 and built the libvirt rpms from source. libvirt-guests was > not running when updating from 2.2.0 to libvirt 2.3.0. I switched > systems and retried on fc23 after rebuilding libvirt 2.2.0 and 2.3.0 > from source. On this system I cannot reproduce the behaviour. > You are correct that the behaviour I got does not match with the > requires directive. So my guess is that my old fc20 might still have > some systemd problem which has been fixed in newer fc versions. That's great to hear! Thanks for providing feedback :) -- Andrea Bolognani / Red Hat / Virtualization -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On 10/07/2016 05:04 PM, Andrea Bolognani wrote: On Fri, 2016-10-07 at 11:42 +0200, Andrea Bolognani wrote: On Fri, 2016-10-07 at 11:19 +0200, Boris Fiuczynski wrote: Andrea, there is another "side effect" of the Requires directive. libvirt-guests gets automatically started when libvirt is updated to v2.3.0. This has some rather nasty implications for the end users. Mh, I don't see why that would happen. The Requires relationship goes in the opposite direction, so if libvirt-guests was not running before the upgrade there should be no reason for it to be started, whether that relationship is there or not. I'll try to reproduce this on my machine and get back to you in a while. I tried upgrading from 2.2.0 to 2.3.0 a bunch of times, but I haven't been able to reproduce the failure you're reporting: libvirt-guests is never started automatically. Can you provide more information? What distribution are you using? Are you building from source, or using your distribution's packages? Are you sure libvirt-guests was not running even before upgrade? -- Andrea Bolognani / Red Hat / Virtualization I used fc20 and built the libvirt rpms from source. libvirt-guests was not running when updating from 2.2.0 to libvirt 2.3.0. I switched systems and retried on fc23 after rebuilding libvirt 2.2.0 and 2.3.0 from source. On this system I cannot reproduce the behaviour. You are correct that the behaviour I got does not match with the requires directive. So my guess is that my old fc20 might still have some systemd problem which has been fixed in newer fc versions. -- Mit freundlichen Grüßen/Kind regards Boris Fiuczynski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martina Köderitz Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294 -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On Fri, 2016-10-07 at 11:42 +0200, Andrea Bolognani wrote: > On Fri, 2016-10-07 at 11:19 +0200, Boris Fiuczynski wrote: > > Andrea, > > there is another "side effect" of the Requires directive. > > libvirt-guests gets automatically started when libvirt is updated to > > v2.3.0. This has some rather nasty implications for the end users. > > Mh, I don't see why that would happen. > > The Requires relationship goes in the opposite direction, > so if libvirt-guests was not running before the upgrade > there should be no reason for it to be started, whether > that relationship is there or not. > > I'll try to reproduce this on my machine and get back to > you in a while. I tried upgrading from 2.2.0 to 2.3.0 a bunch of times, but I haven't been able to reproduce the failure you're reporting: libvirt-guests is never started automatically. Can you provide more information? What distribution are you using? Are you building from source, or using your distribution's packages? Are you sure libvirt-guests was not running even before upgrade? -- Andrea Bolognani / Red Hat / Virtualization -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On Fri, 2016-10-07 at 11:19 +0200, Boris Fiuczynski wrote: > Andrea, > there is another "side effect" of the Requires directive. > libvirt-guests gets automatically started when libvirt is updated to > v2.3.0. This has some rather nasty implications for the end users. Mh, I don't see why that would happen. The Requires relationship goes in the opposite direction, so if libvirt-guests was not running before the upgrade there should be no reason for it to be started, whether that relationship is there or not. I'll try to reproduce this on my machine and get back to you in a while. -- Andrea Bolognani / Red Hat / Virtualization -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On 10/07/2016 09:31 AM, Andrea Bolognani wrote: On Fri, 2016-10-07 at 09:12 +0200, Boris Fiuczynski wrote: Daniel, Andrea, have you looked into the change in behaviour caused by patch 4 of this series? When libvirtd is stopped I get all my running domains suspended. Restarting libvirtd now also causes a disruption since all domains are suspended and than resumed. Before the commit of this patch the domains used to remain running without these disruptions and as a user this is what I expect to happen. I propose to revert commit fb2025ede9ecde1aa04ba6e01ce0c82c9e42dc66. Oh, you're right! Thank you for reporting this issue. We don't want to revert the change though, merely to downgrade the Requires to a Wants. That way, libvirtd stopping (or restarting) will not cause libvirt-guests to stop, but when starting libvirt-guests an attempt will still be made to start up libvirtd if it's not running already. I'll have patches on the list shortly. -- Andrea Bolognani / Red Hat / Virtualization Andrea, there is another "side effect" of the Requires directive. libvirt-guests gets automatically started when libvirt is updated to v2.3.0. This has some rather nasty implications for the end users. -- Mit freundlichen Grüßen/Kind regards Boris Fiuczynski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martina Köderitz Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294 -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On 10/07/2016 09:31 AM, Andrea Bolognani wrote: On Fri, 2016-10-07 at 09:12 +0200, Boris Fiuczynski wrote: Daniel, Andrea, have you looked into the change in behaviour caused by patch 4 of this series? When libvirtd is stopped I get all my running domains suspended. Restarting libvirtd now also causes a disruption since all domains are suspended and than resumed. Before the commit of this patch the domains used to remain running without these disruptions and as a user this is what I expect to happen. I propose to revert commit fb2025ede9ecde1aa04ba6e01ce0c82c9e42dc66. Oh, you're right! Thank you for reporting this issue. We don't want to revert the change though, merely to downgrade the Requires to a Wants. That way, libvirtd stopping (or restarting) will not cause libvirt-guests to stop, but when starting libvirt-guests an attempt will still be made to start up libvirtd if it's not running already. I'll have patches on the list shortly. -- Andrea Bolognani / Red Hat / Virtualization The directive Wants works for me. Just tried it out. -- Mit freundlichen Grüßen/Kind regards Boris Fiuczynski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martina Köderitz Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294 -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On Fri, 2016-10-07 at 09:12 +0200, Boris Fiuczynski wrote: > Daniel, Andrea, > have you looked into the change in behaviour caused by patch 4 of this > series? When libvirtd is stopped I get all my running domains suspended. > Restarting libvirtd now also causes a disruption since all domains are > suspended and than resumed. > Before the commit of this patch the domains used to remain running > without these disruptions and as a user this is what I expect to happen. > I propose to revert commit fb2025ede9ecde1aa04ba6e01ce0c82c9e42dc66. Oh, you're right! Thank you for reporting this issue. We don't want to revert the change though, merely to downgrade the Requires to a Wants. That way, libvirtd stopping (or restarting) will not cause libvirt-guests to stop, but when starting libvirt-guests an attempt will still be made to start up libvirtd if it's not running already. I'll have patches on the list shortly. -- Andrea Bolognani / Red Hat / Virtualization -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
Daniel, Andrea, have you looked into the change in behaviour caused by patch 4 of this series? When libvirtd is stopped I get all my running domains suspended. Restarting libvirtd now also causes a disruption since all domains are suspended and than resumed. Before the commit of this patch the domains used to remain running without these disruptions and as a user this is what I expect to happen. I propose to revert commit fb2025ede9ecde1aa04ba6e01ce0c82c9e42dc66. On 09/06/2016 03:58 PM, Daniel P. Berrange wrote: On Tue, Sep 06, 2016 at 03:55:20PM +0200, Andrea Bolognani wrote: Make libvirt on systemd nicer for the user, by getting rid of some confusing behavior, and overall more solid. More details in each specific patch. ACK to all Regards, Daniel -- Mit freundlichen Grüßen/Kind regards Boris Fiuczynski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martina Köderitz Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294 -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On Tue, 2016-09-06 at 14:58 +0100, Daniel P. Berrange wrote: > On Tue, Sep 06, 2016 at 03:55:20PM +0200, Andrea Bolognani wrote: > > > > Make libvirt on systemd nicer for the user, by getting rid of > > some confusing behavior, and overall more solid. > > > > More details in each specific patch. > > ACK to all Pushed, thanks :) -- Andrea Bolognani / Red Hat / Virtualization -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Re: [libvirt] [PATCH 0/4] systemd-related fixes and improvements
On Tue, Sep 06, 2016 at 03:55:20PM +0200, Andrea Bolognani wrote: > Make libvirt on systemd nicer for the user, by getting rid of > some confusing behavior, and overall more solid. > > More details in each specific patch. ACK to all Regards, Daniel -- |: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :| -- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list