[openstack-dev] [quantum][LBaas] Question about LBaas Extension
Oleg Bondarev
obondarev at mirantis.com
Thu Nov 22 12:09:59 UTC 2012
Hi Leon,
Please see my replies inline
Thanks,
Oleg
From: Leon Cui [mailto:lcui at vmware.com]
Sent: Thursday, November 22, 2012 3:46 PM
To: Oleg Bondarev
Cc: openstack-dev at lists.openstack.org
Subject: [quantum][LBaas] Question about LBaas Extension
Hi Oleg,
I took in your change on LBaas extension and working the CRUD operation
(database access). I met some issues during the implementation. Would you
please take a look?
1. The REST url to "deassociate a health_monitor from a pool" contains
both pool_id and health_monitor_id. By looking at the API defined in
LoadBalancerPluginBase as below, how I got the health_monitor_id? I think
"body" parameter is referring to the request content in JSON rather than the
url parameter. Correct me if I'm wrong.
def delete_health_monitors(self, context, pool_id, body):
[obondarev] I would advise you to freeze the work on health monitors for a
while as it is not decided yet which is the best approach of health monitors
handling API. Most probably it will be changed in my patch soon (pl see '
[openstack-dev] [quantum] [LBaaS] Health monitors REST resource' mail
thread)
2. In ATTRIBUTE_MAP, the "vip_id" attribute of "pools" is validate as
uuid. The question is user can create a pool without specifying the VIP,
which means vip_id is allowed to be empty (None). However, in the
implementation I found that None is unacceptable for type::uuid validator.
The same for "pool_id" attribute of "members".
[obondarev] Good catch, thank you Leon! The 'vip_ip' attribute of a pool
should be validated as 'uuid_or_none'. Regarding 'pool_id' in members it was
decided that it is a required field so that member could be created only for
an existing pool.
3. I think we need to add "pool_id" attribute as part of
"health_monitors". At least set "is_invisible" to False. Otherwise, there
is no way to associate pool and health_monitor?
[obondarev] the same as for 1)
Thanks
Leon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20121122/3986a36f/attachment.html>
More information about the OpenStack-dev
mailing list