Bug 1472306
| Summary: | Horizon overcloud deploy with external load balancer missing 'HTTP_X_FORWARDED_PROTO' configuration. | |||
|---|---|---|---|---|
| Product: | Red Hat OpenStack | Reporter: | Benjamin Schmaus <bschmaus> | |
| Component: | puppet-horizon | Assignee: | Radomir Dopieralski <rdopiera> | |
| Status: | CLOSED WORKSFORME | QA Contact: | nlevinki <nlevinki> | |
| Severity: | low | Docs Contact: | ||
| Priority: | unspecified | |||
| Version: | 10.0 (Newton) | CC: | beth.white, bschmaus, jjoyce, jschluet, rdopiera, slinaber, tvignaud | |
| Target Milestone: | --- | Keywords: | Triaged | |
| Target Release: | 12.0 (Pike) | |||
| Hardware: | x86_64 | |||
| OS: | Linux | |||
| Whiteboard: | ||||
| Fixed In Version: | Doc Type: | If docs needed, set a value | ||
| Doc Text: | Story Points: | --- | ||
| Clone Of: | ||||
| : | 1472887 (view as bug list) | Environment: | ||
| Last Closed: | 2017-07-26 14:57:59 UTC | Type: | Bug | |
| Regression: | --- | Mount Type: | --- | |
| Documentation: | --- | CRM: | ||
| Verified Versions: | Category: | --- | ||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | ||
| Cloudforms Team: | --- | Target Upstream Version: | ||
| Embargoed: | ||||
| Bug Depends On: | ||||
| Bug Blocks: | 1472887 | |||
|
Description
Benjamin Schmaus
2017-07-18 12:27:59 UTC
That configuration line has been removed intentionally, because it prevents the non-https configurations from working. So it is by design then? How does customer achieve the https termination when using a external load balancer then? I don't think anybody actually "designed" this. I think this is just an example of a conflict of interest, where to fix one bug we introduced a regression in other place, and to fix that we regressed the original bug in turn. What is needed is a step back and actual design to make that option be enabled only when it is actually required. I'm not sure yet if this is easily possible right now, but I will find out. I'm sorry, I started looking into this issue, and I realized that I confused it with a different issue, namely the SESSION_COOKIE_SECURE setting. Nevertheless, I'm looking into the puppet-horizon and tripleo-heat-templates files and I can see that the enable_secure_proxy_ssl_header setting that controls this is enabled by default. Can you tell me how was this instance installed? Did you use tripleo/director, or packstack or something else? The customer is using Director for the deployment here. I can't reproduce this problem with OSP12, it seems that it has been fixed in OSP11 judging by this line: https://github.com/openstack/puppet-horizon/blob/stable/ocata/spec/classes/horizon_init_spec.rb#L131 |