Bug 1001128
| Summary: | Unable to install rhq server with rhq.autoinstall.database=overwrite | ||
|---|---|---|---|
| Product: | [Other] RHQ Project | Reporter: | Filip Brychta <fbrychta> |
| Component: | Installer | Assignee: | Stefan Negrea <snegrea> |
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Mike Foley <mfoley> |
| Severity: | high | Docs Contact: | |
| Priority: | high | ||
| Version: | 4.8 | CC: | fbrychta, hrupp |
| Target Milestone: | ER01 | ||
| Target Release: | RHQ 4.9 | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2013-09-24 19:08:38 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
Filip Brychta
2013-08-26 14:28:34 UTC
Filip, is that a brand new install without an existing C* instance? Or an "upgrade" that is trying to re-use the existing C* install by overwriting it? Stefan, I am assigning this over to you. I looked at the drop XML scripts in the schema manager. There is no reason to revert the password of the cassandra user. In fact if anything, we really ought to drop the default cassandra user. The reason we do not delete it goes back to a misunderstanding on my part about bootstrapping C* nodes. At the time I was doing it incorrectly, and while debugging I thought that maybe deleting the cassandra user was causing problems. I also do not think we should be dropping the rhqadmin/internal user. All we should be doing there is dropping the rhq schema. Not sure what all is involved with this issue but I figured you would be better suited to look at it. (In reply to Heiko W. Rupp from comment #1) > Filip, is that a brand new install without an existing C* instance? > Or an "upgrade" that is trying to re-use the existing C* install by > overwriting it? Brand new install without existing C* instance. The issue here was that the drop procedure would fail when the schema was not installed at all. The drop procedure attempts to first re-establish the generic Cassandra user/pass before getting the cluster in a pre-RHQ state. The drop code was wrongly throwing an error if the RHQ user and schema did not exist. The fix catches the first authentication error and takes no action. The drop code will then attempt to establish a Cassandra connection with the generic user name and password and continue the drop procedure. A second authentication is considered an error and will visible to the user. master commit: https://git.fedorahosted.org/cgit/rhq/rhq.git/commit/?id=6eb9a98f7316382cd070922402fc1d9957e41076 Verified on Version: 4.9.0-SNAPSHOT Build Number: a9245fd Bulk closing of RHQ 4.9 verified items |