[Crowbar] [crowbar] openstack swift cluster stays running after removing swift roles from nodes (#663)

Andi_Abes at Dell.com Andi_Abes at Dell.com
Wed May 2 10:14:04 CDT 2012


Dieter, we mostly use the mailing list, rather than github issues....
Not saying it's wrong.. just that's the common take right now.


That said - yes. When you remove a proposal, we don't removing things from nodes. 
Call it a DevOps principle - it's easier to rebuild to a known good state, than try to undo things to get there.
To repurpose the nodes that were previously used for swift you can either (after removing the proposal):
* delete them from crowbar - they'll be rediscovered from scratch
* reinstall them - all the roles the nodes still have will get blasted on the node again (but not swift).

a.

-----Original Message-----
From: Dieter Plaetinck [mailto:reply+i-4385806-bd36dc69f2633af5b247a059df973fa09ab10bf3-782255 at reply.github.com] 
Sent: Wednesday, May 02, 2012 10:30 AM
To: Abes, Andi
Subject: [crowbar] openstack swift cluster stays running after removing swift roles from nodes (#663)

how it is now:
1. create swift proposal, add nodes to it 2. apply it 3. swift cluster is operational and useable 4. remove the proposal, removing roles from the nodes 5. the cluster is still running fine. processes are running, data is still available, cluster can still be used

expected result: 
at least the cluster should go down, swift processes stopped, configuration removed. I would also argue that the data can be removed automatically (it's the crowbar user's job to backup data first if needed, then decomission nodes)

---
Reply to this email directly or view it on GitHub:
https://github.com/dellcloudedge/crowbar/issues/663


More information about the Crowbar mailing list