Bug 1724652 - [RFE] Alow force deletion of in PENDING entities
Summary: [RFE] Alow force deletion of in PENDING entities
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Red Hat OpenStack
Classification: Red Hat
Component: openstack-octavia
Version: 13.0 (Queens)
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
: ---
Assignee: Assaf Muller
QA Contact: Bruna Bonguardo
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2019-06-27 13:24 UTC by Andreas Karis
Modified: 2020-10-26 12:00 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2019-07-03 15:56:01 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Red Hat Knowledge Base (Solution) 4251821 0 None None How to force delete octavia loadbalancer entities that are stuck in PENDING_UPDATE or PENDING_CREATE in Red Hat OpenStac... 2019-06-27 13:34:38 UTC

Description Andreas Karis 2019-06-27 13:24:05 UTC
Description of problem:
Currently, PENDING_UPDATE and PENDING_CREATE entities cannot be deleted. Please allow a force delete, as the only workaround currently is to connect to the database and run e.g.:
~~~
MariaDB [octavia]> update load_balancer set provisioning_status = 'ACTIVE' where provisioning_status = 'PENDING_UPDATE';
Query OK, 1 row affected (0.37 sec)
Rows matched: 1  Changed: 1  Warnings: 0

MariaDB [octavia]> 
~~~

And then delete those entities once set to ACTIVE. 

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


How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info:


Account for the fact that entities may be stuck in PENDING and a force delete is needed.

Comment 1 Gregory Thiemonge 2019-07-03 15:56:01 UTC
Hi Andreas,

Force deletion of LB is not a good idea and is dangerous, it has been discussed upstream in the following story: https://storyboard.openstack.org/#!/story/2002568

We are currently working on solution to mitigate this issue:
- Updating timeout values (users may think a load balancer is stuck while timeouts have not expired yet): https://bugzilla.redhat.com/show_bug.cgi?id=1726733
- Avoid unnecessary service restarts on deployment stack update: https://bugzilla.redhat.com/show_bug.cgi?id=1726735
- Graceful shutdown/restart of octavia containers: https://bugzilla.redhat.com/show_bug.cgi?id=1726738

Comment 2 Andreas Karis 2019-07-03 20:56:15 UTC
I don't think that this is just a timeout issue:
https://bugzilla.redhat.com/show_bug.cgi?id=1712448
https://bugzilla.redhat.com/show_bug.cgi?id=1724798

I ran into these 2 issues above where loadbalancers never left the PENDING state.

Comment 3 Carlos Goncalves 2019-07-04 07:19:53 UTC
RHBZ #1712448 was indeed a bug (already fixed upstream) but it was not leaving resources actually stuck in PENDING_*. It might have given that impression but a closer look showed that there was a bug when an invalid certificate is passed in, the LB was going into ERROR and Octavia after some time tried to repair it.

RHBZ #1724798 does not have resources in a transient state (PENDING_*). I had a quick look at the logs yesterday and I could see them in a non-PENDING status. We will triage it there sometime soon.


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