Bug 1655794 - After upgrading from 4.2 to 4.6, evmserverd service fails to start with rake aborted
Summary: After upgrading from 4.2 to 4.6, evmserverd service fails to start with rake ...
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat CloudForms Management Engine
Classification: Red Hat
Component: Appliance
Version: 5.9.0
Hardware: Unspecified
OS: Unspecified
high
high
Target Milestone: GA
: 5.11.0
Assignee: Joe Rafaniello
QA Contact: Jaroslav Henner
Red Hat CloudForms Documentation
URL:
Whiteboard:
: 1644862 (view as bug list)
Depends On: 1734309
Blocks:
TreeView+ depends on / blocked
 
Reported: 2018-12-03 23:07 UTC by Joe Rafaniello
Modified: 2021-12-10 18:20 UTC (History)
6 users (show)

Fixed In Version: 5.11.0.1
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2019-12-12 13:34:52 UTC
Category: ---
Cloudforms Team: CFME Core
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)
aftermigrate fail evm.log (11.39 KB, text/plain)
2019-07-29 12:27 UTC, Jaroslav Henner
no flags Details


Links
System ID Private Priority Status Summary Last Updated
Red Hat Knowledge Base (Solution) 3669301 0 None None None 2018-12-05 16:07:23 UTC
Red Hat Product Errata RHBA-2019:4199 0 None None None 2019-12-12 13:35:03 UTC

Description Joe Rafaniello 2018-12-03 23:07:11 UTC
Description of problem: The full error is: ActiveRecord::SubclassNotFound: The single-table inheritance mechanism failed to locate the subclass:'ManageIQ::Providers::AnsibleTower::ConfigurationManager::RefreshWorker
It is caused because "ManageIQ::Providers::AnsibleTower::ConfigurationManager::RefreshWorker" was renamed to "ManageIQ::Providers::AnsibleTower::AutomationManager::RefreshWorker" in:
https://github.com/ManageIQ/manageiq/commit/2f0ef1a90758f2b5c37e93a0ed2f586a916dad64

Version-Release number of selected component (if applicable):


How reproducible: Very if you used an ansible tower provider in 4.2 and upgraded to 4.5 or 4.6.


Steps to Reproduce:
1.  Add ansible tower provider in 4.2
2.  Upgrade to 4.6 (or 4.5)
3.  Start evmserverd

Actual results:
It fails to start

Expected results:
It starts.

We triaged and are documenting a workaround for this in bug 1644083.  This new bug is for backporting a fix to remove these rows for any future releases so this workaround isn't required.

Comment 4 Joe Rafaniello 2018-12-05 16:07:24 UTC
*** Bug 1644862 has been marked as a duplicate of this bug. ***

Comment 5 Joe Rafaniello 2018-12-05 16:09:29 UTC
Requesting blocker flag since this will break many upgrades and the backport PRs have been done.  We might not get it backported to all releases due to release schedules but the code is there in case we want to pick it up.

Comment 6 CFME Bot 2018-12-17 17:13:29 UTC
New commit detected on ManageIQ/manageiq-schema/master:

https://github.com/ManageIQ/manageiq-schema/commit/7e1887ef5761b51b369b2af7a52aeaf3ca09b188
commit 7e1887ef5761b51b369b2af7a52aeaf3ca09b188
Author:     Joe Rafaniello <jrafanie>
AuthorDate: Mon Dec  3 17:56:20 2018 -0500
Commit:     Joe Rafaniello <jrafanie>
CommitDate: Mon Dec  3 17:56:20 2018 -0500

    Remove worker rows referencing a renamed worker

    Fixes https://bugzilla.redhat.com/show_bug.cgi?id=1655794

    The worker was renamed in
    https://github.com/ManageIQ/manageiq/commit/2f0ef1a90758f2b5c37e93a0ed2f586a916dad64

 db/migrate/20181203224640_remove_renamed_ansible_tower_configuration_manager_refresh_worker_rows.rb | 10 +
 spec/migrations/20181203224640_remove_renamed_ansible_tower_configuration_manager_refresh_worker_rows_spec.rb | 17 +
 2 files changed, 27 insertions(+)

Comment 7 Jaroslav Henner 2019-07-29 12:26:43 UTC
I restored the db from CFME 5.10 on 5.11 (Currently it is only method I know of to do the upgrade to 5.11), did rake db:migrate and tried to star evmserverd, which failed.

Comment 8 Jaroslav Henner 2019-07-29 12:27:27 UTC
Created attachment 1594254 [details]
aftermigrate fail evm.log

Comment 9 Joe Rafaniello 2019-07-29 15:29:24 UTC
This is a new but similar problem.  This BZ is about miq_workers rows referencing not longer existing classes.  This new issue appears to be a validation on the existing ansible provider rows you had in 5.10.  Please open a new BZ and mark this BZ as blocked by that.  Please provide the test database/appliance ip for that new BZ so we can recreate it and fix it.

Thank you!

Comment 10 Jaroslav Henner 2019-08-06 16:35:22 UTC
On 5.10 I enabled ansible, configured the tower provider, then stopped the evmserverd and dumped the db using pg_dumpall.
Then I copied the keys and other stuff to 5.11 and then restored the db. The evmserverd started fine and the tower is there. Therefore I can conclude the upgrade (although not in-situ) worked.

Comment 12 errata-xmlrpc 2019-12-12 13:34:52 UTC
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-2019:4199


Note You need to log in before you can comment on or make changes to this bug.