Bug 1393337 - [DOCS] [3.4] Improve the OS upgrade solutions
Summary: [DOCS] [3.4] Improve the OS upgrade solutions
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: Documentation
Version: 3.4.0
Hardware: Unspecified
OS: Unspecified
medium
medium
Target Milestone: ---
: ---
Assignee: brice
QA Contact: Anping Li
Vikram Goyal
URL:
Whiteboard:
Depends On:
Blocks: 1383278
TreeView+ depends on / blocked
 
Reported: 2016-11-09 11:10 UTC by Anping Li
Modified: 2017-03-22 03:39 UTC (History)
6 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2017-03-22 03:39:33 UTC
Target Upstream Version:


Attachments (Terms of Use)

Description Anping Li 2016-11-09 11:10:07 UTC
Document URL: 
https://docs.openshift.com/container-platform/3.3/install_config/upgrading/os_upgrades.html#upgrading-os-upgrades-solutions
Section Number and Name: 

Describe the issue: 
The solution is just a highlight

Suggestions for improvement: 
To upgrade OS on Masters
  1) Take a master out of rotation
     oc edit ep/kubernetes -n default
  2) oadm manage-node nodex --schedulable=false
  3) oadm manage-node nodex  --evacuate --force 
  4) upgrade your OS packages  and restart OS
  5) oadm manage-node nodex --schedulable=true

To upgrade OS On nodes. The steps is is following:
  1) oadm manage-node nodex --schedulable=false
  2) oadm manage-node nodex  --evacuate --force 
  3) upgrade your OS packages  and restart OS
  4) oadm manage-node nodex --schedulable=true
Additional information:

Comment 4 Scott Dodson 2017-03-02 20:19:43 UTC
The current process is to 'drain' a node rather than evacuate, the new command is:

oadm drain nodex --force --delete-local-data --ignore-daemonsets

Step #1 for master upgrade is not necessary starting with OCP 3.3 as the endpoint will automatically be removed after an API server has been shut down.

Comment 5 brice 2017-03-03 04:03:24 UTC
Thanks, all.

I've created a PR for this:

https://github.com/openshift/openshift-docs/pull/3865

Anping Li, can I please ask if you think this fulfills this BZ? As the reporter and QE contact, I would also like to ask if the process in the PR is also accurate, as it's changed since your initial report.

Thanks!

Comment 6 Anping Li 2017-03-06 10:40:58 UTC
I think it is better to add description of atomic-openshift-excluder atomic-openshift-docker-excluder which are used avoiding OCP upgrade during OS upgrade.

https://docs.openshift.com/container-platform/3.4/install_config/upgrading/manual_upgrades.html

Comment 7 brice 2017-03-07 00:21:35 UTC
Sure, I've added another step to the procedure.

Can I ask for another look please? Please let me know if the new step should be in a different part of the procedure (if it matters).

Thanks.

Comment 8 Anping Li 2017-03-07 01:51:04 UTC
"Install or update to the following latest available **-excluder* packages ",It should be same version with OCP rather than the latest available . The other part looks good.

Comment 10 brice 2017-03-08 00:41:18 UTC
OK. I've updated the comment as per Comment 8.

Vikram, that's for a different section of the docs, but this could be used for reference for the BZ assignee.

If there's anything else for this BZ, let me know, but otherwise, I'll put this onto peer review.

Thanks, all.

Comment 11 openshift-github-bot 2017-03-09 00:29:39 UTC
Commit pushed to master at https://github.com/openshift/openshift-docs

https://github.com/openshift/openshift-docs/commit/f32df336142847484a4057a6ba251b889b7bad38
Merge pull request #3865 from bfallonf/upgrade_1393337

Bug 1393337 Added procedure to the upgrading OS topic


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