That screams of boot order. It sounds like cloud-init needs to start sooner, like immediately after network initialization. Are these home-grown images?
On Sep 10, 2014, at 11:52 AM, Alex Leonhardt <[email protected]> wrote: > Hi Abel, > > It's actually correct when logging in, but some services, like mcollective > are reporting the wrong fqdn until we reboot the VM so as a temporary fix we > added a reboot at the end of cloud-init. > > Any better way to get this sorted properly though? > > Alex > On 8 Sep 2014 16:20, "Abel Lopez" <[email protected]> wrote: > Yes. I think this is similar to changing hostname on an interactive shell, > you won't see your prompt change until you logout/in. Seems to be a new > feature, maybe something in cloud-init 0.7.5? IIRC it used to be correct at > first login. > > On Monday, September 8, 2014, Alex Leonhardt <[email protected]> wrote: > HI all, > > did anyone have the same problem ? > > Alex > > On 3 September 2014 13:57, Alex Leonhardt <[email protected]> wrote: > Hi, > > I'm using openstack together with puppet and mcollective (VM images is > pre-provisoined) but when a new VM is being spun up, it seems that the host > is being wrongly identified as host-i-p-g-o-e-s-h-e-r-e instead of its actual > hostname that was set by cloud-init / puppet ? > > Am sure there is a trick to make this work properly ? > > Note that after I reboot the VM (no changes made) the host comes up with the > correct name in mcollective. Am just wondering if there is something I forgot > to set in cloud-init or if someone had the same issue and knows a proper fix > (not rebooting) ? > > > Thanks > Alex > > > >
signature.asc
Description: Message signed with OpenPGP using GPGMail
_______________________________________________ OpenStack-operators mailing list [email protected] http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
