In the FS I have proposed we use the 'persistent' flag with the API's. I have 
added this item under 'Open Issues' for now. Will update the FS based on the 
discussion in this list. 

Since a network offering is used by an admin to define the feature set for 
guest networks, now I do think it makes more sense to have the flag in the 
network offering. 
All the default network offerings can have this feature disabled. To create a 
new persistent network a user can use a network offering provided by the admin. 
And to make an existing network persistent, user can update the network 
offering.

Comments?

Thank you,
Likitha

> -----Original Message-----
> From: Chiradeep Vittal [mailto:chiradeep.vit...@citrix.com]
> Sent: Tuesday, January 08, 2013 12:35 PM
> To: CloudStack DeveloperList
> Subject: Re: [DISCUSS] Persistent Networks without a running VM
> 
> We had a discussion on moving the 'persistent' flag to the network offering?
> 
> On 1/7/13 3:38 AM, "Likitha Shetty" <likitha.she...@citrix.com> wrote:
> 
> >Created the first draft of the Functional spec -
> >https://cwiki.apache.org/confluence/display/CLOUDSTACK/FS+-+Persistent+
> >Net
> >works.
> >Will keep updating it based on the feedback.
> >
> >Thank you,
> >Likitha
> >
> >> -----Original Message-----
> >> From: Manan Shah [mailto:manan.s...@citrix.com]
> >> Sent: Friday, January 04, 2013 10:26 PM
> >> To: cloudstack-dev@incubator.apache.org
> >> Subject: Re: [DISCUSS] Persistent Networks without a running VM
> >>
> >> Chiradeep, Likitha,
> >>
> >> My take is that we need to support both kinds of networks (persistent
> >>as well as  non-persistent). Also, I don't think we can have this as a
> >>zone-wide behavior  because not all networks in a zone would need to
> >>be persistent.
> >>
> >> For example, if you are deploying a multi-tier application, you might
> >>only want  the DB tier to be persistent.
> >>
> >> Regards,
> >> Manan Shah
> >>
> >>
> >>
> >>
> >> On 1/3/13 11:31 PM, "Ram Ganesh" <ram.gan...@citrix.com> wrote:
> >>
> >> >Does it make sense to introduce the flag(persistent) as part of
> >> >NetworkOffering?
> >> >
> >> >Thanks,
> >> >RamG
> >> >
> >> >> -----Original Message-----
> >> >> From: Likitha Shetty [mailto:likitha.she...@citrix.com]
> >> >> Sent: 03 January 2013 18:05
> >> >> To: cloudstack-dev@incubator.apache.org
> >> >> Subject: RE: [DISCUSS] Persistent Networks without a running VM
> >> >>
> >> >> Please find my answers and queries inline.
> >> >>
> >> >> Thank you,
> >> >> Likitha
> >> >>
> >> >> > -----Original Message-----
> >> >> > From: Chiradeep Vittal [mailto:chiradeep.vit...@citrix.com]
> >> >> > Sent: Thursday, January 03, 2013 1:03 PM
> >> >> > To: CloudStack DeveloperList
> >> >> > Subject: Re: [DISCUSS] Persistent Networks without a running VM
> >> >> >
> >> >> > So:
> >> >> > 1. There needs to be both kinds of networks available
> >> >> > (persistent as
> >> >> well as non-
> >> >> > persistent) in the same zone?
> >> >> Yes
> >> >>
> >> >> > From an end-user perspective this is going to be confusing since
> >> >> > she
> >> >> has not
> >> >> > been exposed to this internal state before (and generally the
> >> >> > end-
> >> >> user is not
> >> >> > aware of the internal state of the infrastructure).
> >> >> +1. Say we have a new API 'ProvisionNetwork' to provision a
> >> >> +network
> >> >> that has been created by the user. Since the user is not aware of
> >> >> the internal state of a network it would be confusing for the user
> >> >> to understand the difference b/w the 2 API's, CreateNetwork and
> >> >> ProvisionNetwork.
> >> >>
> >> >> > Is it OK to make this behavior
> >> >> > zone-wide, I.e., on every guest network?
> >> >> But this would mean having all networks (in the zone which has
> >> >> this behavior enabled) in an implemented state, even if a network
> >> >> has no physical device or VM deployed in it. This is changing the
> >> >> default CS behavior of not having resources allocated to a network
> >> >> if the network doesn't require it. Is that acceptable ?
> >> >>
> >> >> >
> >> >> >
> >> >> > On 12/31/12 10:19 AM, "Manan Shah" <manan.s...@citrix.com>
> wrote:
> >> >> >
> >> >> > >Thanks Likitha for picking up this requirement. You have
> >> >> > >correctly interpreted the requirements.
> >> >> > >
> >> >> > >Regards,
> >> >> > >Manan Shah
> >> >> > >
> >> >> > >
> >> >> > >
> >> >> > >
> >> >> > >On 12/31/12 2:52 AM, "Likitha Shetty"
> >> >> > ><likitha.she...@citrix.com>
> >> >> wrote:
> >> >> > >
> >> >> > >>Hi,
> >> >> > >>
> >> >> > >>I would like to work on the proposed feature.
> >> >> > >>Restating the requirement. Currently in CloudStack when a user
> >> >> creates
> >> >> > >>a network, a db entry for that network is made, a VLAN ID is
> >> >> assigned
> >> >> > >>and the network is created only when the first VM on that
> >> >> > >>network
> >> >> is
> >> >> > created.
> >> >> > >>With this feature CloudStack should allow users to provision
> >> >> > >>the created network i.e. assign a VLAN ID and implement the
> >> >> > >>network without having to deploy VM's on that network.
> >> >> > >>
> >> >> > >>Comments/Suggestions on the requirement ?
> >> >> > >>
> >> >> > >>Thank you,
> >> >> > >>Likitha
> >> >> > >>
> >> >> > >>-----Original Message-----
> >> >> > >>From: Manan Shah [mailto:manan.s...@citrix.com]
> >> >> > >>Sent: Saturday, December 22, 2012 7:01 AM
> >> >> > >>To: cloudstack-dev@incubator.apache.org
> >> >> > >>Subject: [DISCUSS] Persistent Networks without a running VM
> >> >> > >>
> >> >> > >>Hi,
> >> >> > >>
> >> >> > >>I would like to propose a new feature for persistent networks
> >> >> without
> >> >> > >>running VMs. I have created a JIRA ticket and provided the
> >> >> > >>requirements at the following location.  Please provide
> >> >> > >>feedback on
> >> >> the
> >> >> > requirements.
> >> >> > >>
> >> >> > >>JIRA Ticket:
> >> >> > >>https://issues.apache.org/jira/browse/CLOUDSTACK-706
> >> >> > >>Requirements:
> >> >> >
> >> >> >>https://cwiki.apache.org/confluence/display/CLOUDSTACK/Persisten
> >> >> >>t+N
> >> >> >>et
> >> >> w
> >> >> > >>ork
> >> >> > >>s
> >> >> > >>
> >> >> > >>Regards,
> >> >> > >>Manan Shah
> >> >> > >>
> >> >> > >
> >> >
> >

Reply via email to