Open Stack

Re-sending as plain text..

Hi Anne,

Responses inline.

Kiall

On 10/06/13 14:03, Anne Gentle wrote:
> My Qs:
>
> Is HPCS Hewlett Packard Cloud Services?

Yes - I've updated the application to make this clear :)

>
> Why apply to incubate now, every time I've asked about DNS before it
> sounded like it was a subproject under Networking? What changed the
> perspective on this?

OpenStack Networking, to me, is about moving bits from one point to to 
another. Routing, Load Balancing and VPN all fit this description 
perfectly. DNS on the other hand is about deciding where to send those bits.

As a user of $cloud's DNS service, I would expect the DNS service to 
allow me to create any domain/zone I want, and direct records to IPs 
which are outside the cloud's control e.g. your existing datacenters 
etc. These ideas don't seem to have any parallels inside OpenStack 
Networking.

So - Why now? At the Portland summit, we gave a talk about Moniker. We 
had quite a few people asking when we would apply. So - the HPCS DNS 
team, Ryan from Wikimedia and JC from from eBay got together and wrote 
the application then and there.

>
> What other pluggable drivers are possible for DNS (and how many exist
> now)? I just see Ryan's planned driver.

Currently, we have great support for PowerDNS, good support for BIND9 
and a few other mostly toy implementations. At HP, we use a combination 
of the PowerDNS driver and a Akamai driver (Akamai's DNS API is under 
NDA, so we can't publish the implementation).

>
> Do you integrate with metering already or is it in your plan?

We emit (most) of the necessary events for Metering to consume, but have 
not started any integration.

>
> Looks like you still have name changing to do for the documentation.

Yup - Lots of rename still to happen. We're holding off on renaming the 
repo's etc until this application is accepted/rejected on the advice 
Monty, as the rename involves Gerrit downtime.. So, if we rename now and 
then get accepted, Gerrit would need 2 separate downtimes..

>
> How updated is the DNS solution comparison document referred to on the
> wiki page?

The comparison document was is fairly outdated. It was written pre-San 
Diego summit to help with discussing which implementation, if any, 
should be pursued as "official".

Thanks,
Kiall