Bug 1245572
| Summary: | overcloud: HA deployment: Glance api.log flooded by warning messages. | ||
|---|---|---|---|
| Product: | Red Hat OpenStack | Reporter: | Leonid Natapov <lnatapov> |
| Component: | openstack-puppet-modules | Assignee: | Giulio Fidente <gfidente> |
| Status: | CLOSED ERRATA | QA Contact: | Leonid Natapov <lnatapov> |
| Severity: | high | Docs Contact: | |
| Priority: | high | ||
| Version: | Director | CC: | eglynn, fdinitto, fpercoco, gfidente, lbezdick, lnatapov, mburns, rbiba, rohara, yeylon |
| Target Milestone: | z2 | Keywords: | Triaged, ZStream |
| Target Release: | 7.0 (Kilo) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | openstack-puppet-modules-2015.1.8-14.el7ost | Doc Type: | Bug Fix |
| Doc Text: |
Previously, the HAProxy configuration contained the "option httpchk GET /" directive. Consequently, a "GET /" request was sent every two seconds, and the following message was repeated indefinitely in the /var/log/glance/api.log file:
WARNING glance.api.middleware.version_negotiation [req-...] Unknown version. Returning version choices.
This update removes the aforementioned directive. As a result, the warning message is no longer logged.
|
Story Points: | --- |
| Clone Of: | Environment: | ||
| Last Closed: | 2015-10-08 12:23:37 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: | |||
|
Description
Leonid Natapov
2015-07-22 10:37:20 UTC
*** Bug 1245573 has been marked as a duplicate of this bug. *** Leonid, Would you mind sharing what you did to get all these messages? The message is logged when someone tries to call a glance endpoint with a non-valid version number. I don't see anything wrong with this log. Please, provide - always - as much information as possible to understand the issue. This report, as it is, is not useful at all. Who is calling Glance? Why is it being called with the wrong version number? How was glance configured? etc. glance configured with iscsi backend. I have downloaded cirros images and created several VMs. That's all I did. I don't know why it's being called with the wrong number. Things seems to be working ok. Image was downloaded and I created several instances using that image. I just wanted to check something in glance log and when I opened it I saw this flood of messages. Happen with ceph as well. Flavio,I have an environment if you want to look at the logs. I will send you an email with the environment details. I looked into the box. Apparently there's a process doing periodic request's to Glance and that process is not passing the version as part of the URL. What are you using for HA? Is any of these tools continuously pinging Glance? What do you mean by "what are you using for HA" ?. Pls,let's talk on IRC in real time. My nickname is lesik. I dug into this myself. You've haproxy installed and it's configured to run `GET` requests on glance-api to know if it's alive. The requests don't contain a version in hte URL, which is causing all these log lines. This is your haproxy config: listen glance_api bind 192.0.2.2:9292 bind 192.0.2.3:9292 option httpchk GET / server 192.0.2.1 192.0.2.1:9292 check fall 5 inter 2000 rise 2 listen glance_registry bind 192.0.2.3:9191 server 192.0.2.1 192.0.2.1:9191 check fall 5 inter 2000 rise 2 I'd recommend using `/versions` instead of `/` Fabio, do you know who to ping to change `haproxy`'s official configuration and docs? (In reply to Flavio Percoco from comment #10) > Fabio, > > do you know who to ping to change `haproxy`'s official configuration and > docs? I would talk to Ryan O´hara for the configuration (to check what is the best config) and Giulio Fidente to make the change in OSPd. For the doc, maybe Summer Long, but I am not entirely sure. Ryan, Giulio Mind taking a look at this? Feel free to move the bug respectively. Probably due to the haproxy httpchk option currently configured to send a 'GET /' every 2 secs. Should go away with: https://review.openstack.org/#/c/197904 (In reply to Flavio Percoco from comment #12) > Ryan, Giulio > > Mind taking a look at this? Feel free to move the bug respectively. I don't think you want httpchk option, but that may have changed. I know there was an effort upstream to get httpchk working with the various API services, but I do not know the current state of this effort. It looks like Guilio is correct in comment #13. No flood in the api.log openstack-puppet-modules-2015.1.8-17.el7ost.noarch Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2015:1872 |