Hi Didi, sorry for the delay in reporting final results. I can confirm that simply creating /etc/pki/libvirt suffices: the "private" subdir gets automatically correctly created (along with all certificates etc.) and setup completes fine.
It did not start up because of BZ #1075126 (HA agent died) but I noticed (by chance) that there is an updated (18/03/2014) workaround and after manually applying it (no oVirt new packages have been published yet) the Engine VM started fine. Now I'm battling with an (apparently) GlusterFS (3.5.0beta4) bug that makes the (NFS based, but Gluster-provided) Engine storage domain shutdown by itself after a while (causing Engine VM to die). Many thanks again for your support, Giuseppe PS: would you suggest a complete reinstall with GlusterFS 3.4.x stable instead? ;> PS2: sorry for top-posting (but Hotmail keeps failing on proper quoting...) From: [email protected] To: [email protected] Date: Sun, 16 Mar 2014 15:14:07 +0100 CC: [email protected] Subject: Re: [Users] Self-hosted-engine setup error Date: Sun, 16 Mar 2014 05:14:39 -0400 From: [email protected] To: [email protected] CC: [email protected] Subject: Re: [Users] Self-hosted-engine setup error From: "Giuseppe Ragusa" <[email protected]> To: "Yedidyah Bar David" <[email protected]> Cc: [email protected] Sent: Saturday, March 15, 2014 2:15:18 AM Subject: RE: [Users] Self-hosted-engine setup error Hi Didi, Date: Thu, 13 Mar 2014 02:46:50 -0400 From: [email protected] To: [email protected] CC: [email protected] Subject: Re: [Users] Self-hosted-engine setup error From: "Giuseppe Ragusa" <[email protected]> To: [email protected] Sent: Thursday, March 13, 2014 2:38:42 AM Subject: [Users] Self-hosted-engine setup error Hi all, while attempting a from-scratch self-hosted-engine installation on CentOS 6.5 (also freshly reinstalled from scratch) on a physical node (oVirt 3.4.0_pre + GlusterFS 3.5.0beta4; NFS storage for engine VM), the process fails almost immediately with: [root@cluster1 ~]# ovirt-hosted-engine-setup --config-append=/root/ovhe-setup-answers.conf [ INFO ] Stage: Initializing Continuing will configure this host for serving as hypervisor and create a VM where you have to install oVirt Engine afterwards. Are you sure you want to continue? (Yes, No)[Yes]: [ INFO ] Generating a temporary VNC password. [ INFO ] Stage: Environment setup Configuration files: ['/root/ovhe-setup-answers.conf'] Log file: /var/log/ovirt-hosted-engine-setup/ovirt-hosted-engine-setup-20140313010526.log Version: otopi-1.2.0_rc3 (otopi-1.2.0-0.9.rc3.el6) [ INFO ] Hardware supports virtualization [ INFO ] Bridge ovirtmgmt already created [ INFO ] Stage: Environment packages setup [ INFO ] Stage: Programs detection [ INFO ] Stage: Environment setup [ INFO ] Generating VDSM certificates [ ERROR ] Failed to execute stage 'Environment setup': [Errno 2] No such file or directory: '/etc/pki/libvirt/clientcert.pem' I already got another such report yesterday - seems like a bug in the fix for https://bugzilla.redhat.com/show_bug.cgi?id=1034634 .I hope to push a fix later today. I look forward to have the fix pushed/merged in actual packages. [ INFO ] Stage: Clean up [ INFO ] Stage: Pre-termination [ INFO ] Stage: Termination The /root/ovhe-setup-answers.conf has been saved from a previous installation (before reinstalling) and only minimally edited (removed some lines with UUIDs etc.). The /etc/pki/libvirt dir is completely missing on both nodes; last time I tried the whole setup I do not recall of having such problems, but maybe something was different then. The generated /var/log/ovirt-hosted-engine-setup/ovirt-hosted-engine-setup-20140313010526.log has been saved as: http://pastebin.com/ezAJETBN I hope to be able to progress further to test the whole 2-nodes setup (second node freshly reinstalled too and already up with GlusterFS and waiting to be added to oVirt cluster) and datacenter configuration. Many thanks in advance for any suggestions/help, For now, you can simply:mkdir /etc/pki/libvirt This should be enough. The workaround works: the self-hosted-engine installation proceeds now. Thanks for the report!-- Didi Many thanks for your kind and prompt assistance, Giuseppe Thanks for the report. The workaround is probably not enough, depends on what youare trying to do. 'mkdir /etc/pki/libvirt/private' is needed too. Without it, the code thatcopies there a key will create a file 'private' instead of copying it into a directory 'private'. Fix [1] was merged to all branches. [1] http://gerrit.ovirt.org/25747 Best regards,-- Didi Hi Didi, the workaround seemed actually to be enough to make the self-hosted-engine setup go through up to the end without any user-visible error, but it left me with a non-running Engine VM afterwards (basically it did not restart up automatically under HA protection). I collected all logs and reported it in a separate message with subject "Self-hosted engine setup ok but no engine vm running‏ " but got no comments yet. If I get no suggestions I will try to perform some "corrective actions" based on my understanding of the problems at hand, but I did not want to "corrupt" the exact state to help in debugging (starting again from scratch with full first node reinstallation is a somewhat lenghty process since I have no local repo mirrors). Many thanks again, Giuseppe _______________________________________________ Users mailing list [email protected] http://lists.ovirt.org/mailman/listinfo/users
_______________________________________________ Users mailing list [email protected] http://lists.ovirt.org/mailman/listinfo/users
