Bug 1583064
| Summary: | Should not launch multi unbind sandboxes frequently while unbind failed | ||
|---|---|---|---|
| Product: | OpenShift Container Platform | Reporter: | Zhang Cheng <chezhang> |
| Component: | Service Broker | Assignee: | Jesus M. Rodriguez <jesusr> |
| Status: | CLOSED ERRATA | QA Contact: | Zhang Cheng <chezhang> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 3.10.0 | CC: | aos-bugs, jiazha, jmatthew, zhsun, zitang |
| Target Milestone: | --- | ||
| Target Release: | 3.10.0 | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: |
Cause:
The Service Catalog changed how it handles failed bind and unbind actions, by repeatedly calling the action for a long period of time. APBs missing bind or unbind actions will cause binds or unbind actions to fail.
Consequence:
Because the Service Catalog detects a failure, repeatedly calls the Ansible Service Broker which will create new projects for each of the APBs it is launching in response to the action.
Fix:
The Ansible Service Broker will detect if the action is missing and mark the job as succeeded and return an error string in the message.
Result:
Only one namespace is created for a missing bind or unbind action playbook. If the bind or unbind *exists* and fails, this fix does not affect that we expect that to fail requiring APB authors to fix their APBs.
|
Story Points: | --- |
| Clone Of: | Environment: | ||
| Last Closed: | 2018-07-30 19:16:18 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
Zhang Cheng
2018-05-28 07:59:08 UTC
If the unbind fails we really want to let them know it failed because it is probably something the APB developer needs to fix. But if the action is non-existent which is the case with this bug, we really need to prevent the multiple sandboxes. release-1.2 PR: https://github.com/openshift/ansible-service-broker/pull/971 master PR: https://github.com/openshift/ansible-service-broker/pull/972 $ oc get projects NAME DISPLAY NAME STATUS ansible-service-broker Active blog-project Active default Active dh-postgresql-apb-bind-zc82n Active dh-postgresql-apb-prov-jq97t Active dh-postgresql-apb-unbi-vpn2r Active Looking at the asb log BEFORE this fix: $ grep UNBINDING multi-unbind-asb.log time="2018-05-30T17:31:07Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:11Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:16Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:21Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:26Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:31Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:40Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:44Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:49Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:55Z" level=info msg=" UNBINDING " time="2018-05-30T17:31:59Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:04Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:08Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:13Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:17Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:23Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:29Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:33Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:38Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:47Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:51Z" level=info msg=" UNBINDING " time="2018-05-30T17:32:56Z" level=info msg=" UNBINDING " time="2018-05-30T17:33:03Z" level=info msg=" UNBINDING " time="2018-05-30T17:33:08Z" level=info msg=" UNBINDING " time="2018-05-30T17:33:12Z" level=info msg=" UNBINDING " Looking at the asb log AFTER this fix: $ grep UNBINDING asb.log time="2018-05-30T20:03:10Z" level=info msg=" UNBINDING " https://errata.devel.redhat.com/advisory/33505 updated with the following builds: openshift-enterprise-apb-tools-v3.10.0-0.32.0.3 openshift-enterprise-asb-container-v3.10.0-0.54.0.1 openshift-enterprise-mariadb-apb-v3.10.0-0.51.0.1 openshift-enterprise-mediawiki-apb-v3.10.0-0.54.0.1 openshift-enterprise-mediawiki-container-v3.10.0-0.54.0.0 openshift-enterprise-mysql-apb-v3.10.0-0.54.0.1 openshift-enterprise-postgresql-apb-v3.10.0-0.54.0.1 Verified and passed with asb:1.2.16 Follow original steps, sandbox of unbind will be automated trigger while bind failed, and be prevented the multiple sandboxes in currently. LGTM. # oc get ns | grep postgre rh-postgresql-apb-bind-ml4qw Active 11m rh-postgresql-apb-prov-k25kz Active 13m rh-postgresql-apb-unbi-6pzvb Active 10m # oc logs asb-2-89lg4 | grep UNBINDING time="2018-05-31T07:09:17Z" level=info msg=" UNBINDING " 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-2018:1816 |