Loading...

XML

Word

Printable

Details

Type: Bug
Resolution: Unresolved
Priority: Minor
Fix Version/s: 4.14.0
Affects Version/s: 4.9
Component/s: Installer / OpenShift on OpenStack
Labels:

Test Coverage:

-
Severity:
Moderate
Sprint:
ShiftStack Sprint 233, ShiftStack Sprint 234, ShiftStack Sprint 235, ShiftStack Sprint 236
sprint_count:
4
Architecture:

Unspecified
Release Note Text:

Hide
In the deprovision phase, the Installer deletes all Object storage containers linked to the OCP cluster. In order to delete the containers, all objects in them must be deleted.
Before this patch, bulk object deletion was initiated sequentially, in batches of 50 objects. With this change, bulk object deletion is executed with a maximum of 3 concurrent calls to the OpenStack API, each deleting an unbounded number of objects, based on the default pagination of the object listing endpoint (which accounts to 10000 in the default OpenStack settings).

Show
In the deprovision phase, the Installer deletes all Object storage containers linked to the OCP cluster. In order to delete the containers, all objects in them must be deleted. Before this patch, bulk object deletion was initiated sequentially, in batches of 50 objects. With this change, bulk object deletion is executed with a maximum of 3 concurrent calls to the OpenStack API, each deleting an unbounded number of objects, based on the default pagination of the object listing endpoint (which accounts to 10000 in the default OpenStack settings).
Release Note Type:
Bug Fix

SFDC Cases Counter:
SFDC Cases Links:

Description

Version:

$ openshift-install version

./openshift-install 4.9.11
built from commit 4ee186bb88bf6aeef8ccffd0b5d4e98e9ddd895f
release image quay.io/openshift-release-dev/ocp-release@sha256:0f72e150329db15279a1aeda1286c9495258a4892bc5bf1bf5bb89942cd432de
release architecture amd64

Platform: Openstack

install type: IPI

What happened?

Image streams using the swift container to store the images, after running so many image streams I am able to see the huge number of objects in the swift container if I destroy the cluster now, it takes huge time based on the size of the swift container

What did you expect to happen?

The destroy script should clean the resources in some reasonable time

How to reproduce it (as minimally and precisely as possible)?

deploy OCP, run some workload which creates a lot of image streams and destroy the cluster, it will take a lot of time to complete the destroy cmd

Anything else we need to know?

here is the output of the swift state cmd and the time it took to complete the destroy job

$ swift stat vlan609-26jxm-image-registry-nseyclolgfgxoaiysrlejlhvoklcawbxt
Account: AUTH_2b4d979a2a9e4cf88b2509e9c5e0e232
Container: vlan609-26jxm-image-registry-nseyclolgfgxoaiysrlejlhvoklcawbxt
Objects: 723756
Bytes: 652448740473
Read ACL:
Write ACL:
Sync To:
Sync Key:
Meta Name: vlan609-26jxm-image-registry-nseyclolgfgxoaiysrlejlhvoklcawbxt
Meta Openshiftclusterid: vlan609-26jxm
Content-Type: application/json; charset=utf-8
X-Timestamp: 1640248399.77606
Last-Modified: Thu, 23 Dec 2021 08:34:48 GMT
Accept-Ranges: bytes
X-Storage-Policy: Policy-0
X-Trans-Id: txb0717d5198e344a5a095d-0061c93b70
X-Openstack-Request-Id: txb0717d5198e344a5a095d-0061c93b70