I think because I ran the 'osdeploy initialize -i' with just 'sudo' it didn't exit cleanly and I just ignored it, which means it wasn't really completely setup properly from the beginning. I'm used to sudo being a go/no-go situation, not a partial success. So had I run it with sudo -i or maybe sudo sh -c 'blah ...' or something, that would have worked. Once I reran it as root the exit was clean, and the selfcheck also exited cleanly. Lesson learned. Thanks for the quick response, I'm looking forward to the next steps of configuration with Ansible (as opposed to postscripts).
Brian Joiner On Tue, Feb 13, 2024 at 3:47 PM Jarrod Johnson <jjohns...@lenovo.com> wrote: > Well, I'll play with sudo a bit because that was... confusing... > > And really not sure why it was stuck at that phase if selfcheck didn't say > anything but now it works... > > Certainly not a waste of time, just wish I knew better how it happened. > Glad it worked though. > ------------------------------ > *From:* Brian Joiner <martinitime1...@gmail.com> > *Sent:* Tuesday, February 13, 2024 3:48 PM > *To:* xCAT Users Mailing list <xcat-user@lists.sourceforge.net> > *Subject:* Re: [xcat-user] [External] Confluent PXE fail API Token check > deployment.apiarmed > > Embarrassing. Ran as the true root user instead of sudo and the osdeploy > ran with no errors, as did the self check. And now it's deploying because > of course it is. > > Well sorry for wasting everyone's time lol > > On Tue, Feb 13, 2024, 2:05 PM Brian Joiner <martinitime1...@gmail.com> > wrote: > > Output: dbm.error: db file doesn't exist; use c or n flag to create a > new db > > Apologies, I may have gotten some errors on the 'osdelply initialize' > stage an forgot, so I ran it again: > > FileNotFoundError: [Erro 2] No such file or directory: 'collective' > > Brian Joiner > > > On Tue, Feb 13, 2024 at 12:15 PM Jarrod Johnson <jjohns...@lenovo.com> > wrote: > > confluent_selfcheck -n <node> > > Does it report anything of interest? > ------------------------------ > *From:* Brian Joiner <martinitime1...@gmail.com> > *Sent:* Tuesday, February 13, 2024 12:38 PM > *To:* xCAT Users Mailing list <xcat-user@lists.sourceforge.net> > *Subject:* [External] [xcat-user] Confluent PXE fail API Token check > deployment.apiarmed > > I have a confluent system running in my vmware lab at home, and after some > time I got it to deploy VM's successfully by disabling the firewall on > the confluent head node. I use FreeIPA for identity and DNS (which was a > huge help getting going with Confluent, I highly recommend it). > > I'm replacing the xCAT head node at work with the same setup (Confluent on > Rocky 9.3, FreeIPA id server handling DNS for the management network and > domain). FreeIPA works great, sudo rules work as expected, etc. > > However, even with SElinux disabled and firewalld stopped nodes cannot get > past the first few boot steps. PXE works, the node gets the right > ip/dns/etc, but the process fails with: > > dracut-pre-trigger Failed getting API token, check > deployment.apiarmed attribute on <node> > > The attribute looks right on the node (once). > > I've followed the quick start and other guides (as far as I know) but I > can't get past this to continue deployment. Confluent event log shows PXE > offer to the node, and nothing else. As I mentioned, Confluent works fine > at home as long as the firewall is disabled (still haven't found the right > combo of rules to fix it) but does not work in my workplace env. Is there > a way to get more verbose logging to see what's at issue? > > > Brian Joiner > _______________________________________________ > xCAT-user mailing list > xCAT-user@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/xcat-user > > _______________________________________________ > xCAT-user mailing list > xCAT-user@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/xcat-user >
_______________________________________________ xCAT-user mailing list xCAT-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xcat-user
