Bug 1514487 - The installer should delete etcd2 kube data if platform is using etcd3
Summary: The installer should delete etcd2 kube data if platform is using etcd3
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: Installer
Version: 3.7.0
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
: 3.11.z
Assignee: Vadim Rutkovsky
QA Contact: ge liu
URL:
Whiteboard:
: 1541550 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2017-11-17 14:30 UTC by Justin Pierce
Modified: 2019-01-10 09:04 UTC (History)
9 users (show)

Fixed In Version:
Doc Type: Enhancement
Doc Text:
Feature: A new playbook was added to cleanup etcd2 data Reason: If the cluster was upgraded from 3.5 it might still carry etcd2 data and use up space. Result: A new playbook was added to safely remove etcd2 data
Clone Of:
Environment:
Last Closed: 2019-01-10 09:03:57 UTC
Target Upstream Version:

Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Red Hat Product Errata RHBA-2019:0024 0 None None None 2019-01-10 09:04:04 UTC

Description Justin Pierce 2017-11-17 14:30:01 UTC 
Description of problem:
Even if there is a successful migration to etcdv3, etcd2 keys are left behind in the etcd database. It has been recommended that this data be removed to conserve IO & memory use. 

Version-Release number of the following components:
v3.7.9
Comment 5 Scott Dodson 2018-08-23 20:43:51 UTC 
*** Bug 1541550 has been marked as a duplicate of this bug. ***
Comment 6 Stephen Cuppett 2018-08-24 13:04:55 UTC 
If we don't add it on the current release we already would have missed it for people which have upgraded past 3.9 to 3.10. Is the etcdctl2 command still present in 3.11 for us to use?

We can either try and add this for 3.11 to automatically catch the cleanup and/or clone a doc bug for 3.7 docs to provide cleanup guidance for those who attempt to find it.
Comment 7 Eric Rich 2018-09-04 02:26:49 UTC 
Why not make this an optional release note (docs update) in 3.7+ and with 3.11 automate the task?
Comment 8 Vadim Rutkovsky 2018-09-04 12:26:38 UTC 
Created https://github.com/openshift/openshift-ansible/pull/9899 to add a playbook to clean up the existing data
Comment 9 Vadim Rutkovsky 2018-09-17 07:50:19 UTC 
Fix is available in openshift-ansible-3.11.0-0.33.0
Comment 13 Vadim Rutkovsky 2018-09-21 09:27:47 UTC 
Right, can reproduce. `failed_when` didn't work as I expected it to.

PR to master branch - https://github.com/openshift/openshift-ansible/pull/10184
Comment 14 Vadim Rutkovsky 2018-09-24 10:34:46 UTC 
3.11 release cherrypick - https://github.com/openshift/openshift-ansible/pull/10199
Comment 15 Vadim Rutkovsky 2018-10-19 08:52:32 UTC 
Fix is available in openshift-ansible-3.11.26-1
Comment 16 ge liu 2018-10-23 07:42:14 UTC 
Verified with openshift-ansible-3.11.28-1

1).# etcdctl2 mkdir /kubernetes.io/events/default
# etcdctl2 get /kubernetes.io
/kubernetes.io: is a directory

2). Run playbook: /usr/share/ansible/openshift-ansible/playbooks/openshift-etcd/remove-etcdv2-data.yml successfully.

3). Check the v2 data:
# etcdctl2 get /kubernetes.io
migrated
Comment 18 errata-xmlrpc 2019-01-10 09:03:57 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-2019:0024
Note You need to log in before you can comment on or make changes to this bug.