Bug 1293603
| Summary: | [RFE] [hosed-engine-setup] Allow setting engine's FQDN only when it's actually needed. | ||
|---|---|---|---|
| Product: | [oVirt] ovirt-hosted-engine-setup | Reporter: | Nikolai Sednev <nsednev> |
| Component: | RFEs | Assignee: | Sandro Bonazzola <sbonazzo> |
| Status: | CLOSED NOTABUG | QA Contact: | Nikolai Sednev <nsednev> |
| Severity: | low | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 1.3.1.3 | CC: | bugs, didi, gklein, lveyde, rmartins, sbonazzo, stirabos |
| Target Milestone: | --- | Keywords: | FutureFeature |
| Target Release: | --- | Flags: | rule-engine:
planning_ack?
rule-engine: devel_ack? rule-engine: testing_ack? |
| Hardware: | x86_64 | ||
| OS: | Linux | ||
| URL: | https://bugzilla.redhat.com/show_bug.cgi?id=1099911 | ||
| Whiteboard: | integration | ||
| Fixed In Version: | Doc Type: | Enhancement | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2015-12-23 10:14:44 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | Integration | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
Nikolai Sednev
2015-12-22 11:18:38 UTC
we're not going to move the configuration of hosted engine out of the customization stage of OTOPI. In your procedure: > Steps to Reproduce: > 1.ovirt-hosted-engine --deploy and then install via pxe rhel6.5 and yum update -y to latest and yum install rhevm. Here you miss to set the FQDN of the provisioned VM with the FQDN you set during configuration stage of hosted-engine. > 2.rhevm-setup and get running engine. > 3.enter to UI via given IP address of engine. > 4.continue with the ovirt-hosted-engine deploy, while receiving "[ ERROR ] Sandro, I think this bug is useful to fix. At least in our environment, and I suspect in many others, there are systems set up so that you can automatically install an OS from PXE, get from dhcp both a name and IP address, and this name will already be setup in the DNS to point at the IP address. If you want to rely on such a system to ease your dev/qe/whatever work, and do not mind much about the actual name you'll get based on what address dhcpd decided to give you, you want to be able to provide the name you actually got after OS install at that point, and not before you started the OS install. Obviously not high priority, and I agree it breaks our current notion of host hosted-engine deploy should work, but we already do ask questions at such late stages - e.g. re the DC (or cluster) name to add the engine to. So I'd say it's not a drastic change. If you want to use different naming than the one provided by DNS/DHCP you still can change /etc/hosts (run dnsmasq locally so it will load from /etc/hosts and change resolv.conf to use that), no? (In reply to Sandro Bonazzola from comment #3) > If you want to use different naming than the one provided by DNS/DHCP you > still can change /etc/hosts (run dnsmasq locally so it will load from > /etc/hosts and change resolv.conf to use that), no? Yes, but that's not the point. What I personally do is enter the fqdn I want to use, and add this name to the local /etc/hosts with a fake IP address. After I finish installing the OS, I check what ip address it got from dhcp, and put this in /etc/hosts. I guess that's a good enough workaround, so leaving closed. |