Send kea-dev mailing list submissions to
[email protected]
To subscribe or unsubscribe via the World Wide Web, visit
https://lists.isc.org/mailman/listinfo/kea-dev
or, via email, send a message with subject or body 'help' to
[email protected]
You can reach the person managing the list at
[email protected]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of kea-dev digest..."
Today's Topics:
1. Re: Design for Kea Host Reservation (Angelo Failla)
2. Re: Design for Kea Host Reservation (Marcin Siodelski)
----------------------------------------------------------------------
Message: 1
Date: Tue, 30 Sep 2014 12:16:20 +0000
From: Angelo Failla <[email protected]>
To: Marcin Siodelski <[email protected]>, Kea Dev List
<[email protected]>
Subject: Re: [kea-dev] Design for Kea Host Reservation
Message-ID: <d050683e.1d0e2%[email protected]>
Content-Type: text/plain; charset="iso-8859-1"
This looks promising, I still have to read the doc properly but here is
one question for you :)
Are you guys thinking of making this new HR code extensible in way similar
to the Hook APIs?
We currently use bind10 in production with a custom hook lib that bypasses
subnet allocation logic and just fetches this info connecting to one of
our backend systems.
It would be very powerful to make people able to override the source class
(HostDataSource) or just provide more hook points (I believe this is how
you call them) so that you could fetch host information from other sources
that are not mysql databases or any supported source (think about some
backend system already present in people?s production networks)
I spoke with Vicky a while ago promising that I?ll be working on testing
my hook code with KEA0.9, unfortunately I didn?t manage to get to that
point, we are still using bind10, but we should migrate to KEA0.9 starting
end of October.
--
Angelo Failla
Facebook Cluster Ops - Dublin
[email protected]
On 9/30/14, 1:20 PM, "Marcin Siodelski" <[email protected]> wrote:
>All,
>
>I have created a new design document on Kea wiki:
>https://urldefense.proofpoint.com/v1/url?u=http://kea.isc.org/wiki/HostRes
>ervationDesign&k=ZVNjlDMF0FElm4dQtryO4A%3D%3D%0A&r=ScT1tCv0IwhfW9O3zK6EHw%
>3D%3D%0A&m=h7aynHbecCkIIRMjTjqkKdfELJO4akpcCBJAJ8lxsEk%3D%0A&s=9c6d625fa62
>f63fe5112a36362acfcd75892e58c6921e4b9cae9a0b7035805b4 for Host
>Reservation. It
>summarizes the discussions that we conducted on a mailing list and
>during the Kea hackathon.
>
>The document introduces detailed layout of the database for host
>reservations. It also presents relations between the old and new C++
>classes and some other details.
>
>At present, the document doesn't cover the design of a management tool
>for updating and adding new host reservations to the database. I think
>this tool is out of scope for now.
>
>The Trac ticket for creating a HR design is #3559. This ticket is now in
>the review queue and the regular review should be conducted by one of
>ISC engineers. I also encourage mail list subscribers to have a look and
>comment.
>
>Marcin
>_______________________________________________
>kea-dev mailing list
>[email protected]
>https://urldefense.proofpoint.com/v1/url?u=https://lists.isc.org/mailman/l
>istinfo/kea-dev&k=ZVNjlDMF0FElm4dQtryO4A%3D%3D%0A&r=ScT1tCv0IwhfW9O3zK6EHw
>%3D%3D%0A&m=h7aynHbecCkIIRMjTjqkKdfELJO4akpcCBJAJ8lxsEk%3D%0A&s=288fd79e92
>4c7fcf3175bd321661e7ef6c84ea16967ac4b775b5c263a461fceb
------------------------------
Message: 2
Date: Tue, 30 Sep 2014 15:11:40 +0200
From: Marcin Siodelski <[email protected]>
To: Angelo Failla <[email protected]>, Kea Dev List
<[email protected]>
Subject: Re: [kea-dev] Design for Kea Host Reservation
Message-ID: <[email protected]>
Content-Type: text/plain; charset=windows-1252
Angelo,
The HostMgr class is meant to use any derivation of HostDataSource or
CachedHostDataSource instead of MySQL, or any other backend we will
implement in the future. This is a similar concept as the one that we
implemented for lease storage.
In other words, if you have a proprietary storage for your host
reservations (which maybe holds some additional information about
hosts/subscribers connected to your network), you'll need to implement a
class that derives from HostDataSource or CachingHostDataSource which
will retrieve the required information from your proprietary storage and
setup the HostMgr to use this data source, instead of MySQL.
Having said that, it would be quite interesting to get the feedback from
you about the proposed API that the HostDataSource exposes. In
particular, it would like to know if the amount and type of data is
sufficient and that the criteria for searching for host reservations is
sufficient (MAC, DUID etc.)
The HostMgr will be a front end for the allocation engine to obtain a
reservation for the host before it attempts to dynamically allocate a
lease. Regardless, whether the allocated prefix or address comes from
the reservation (via HostMgr) or is dynamically allocated, you will
still have a chance to override this selection via hooks.
You will still be able bypass the subnet selection step with hooks, but
the HostDataSource API assumes existence of some subnet id to identify
which subnet the client is connected to. Obviously, if your custom
libraries don't make use of subnet id, your custom backend may ignore
subnet id value passed via API and the dummy value can be passed to
getHost functions.
Marcin
On 30/09/14 14:16, Angelo Failla wrote:
> This looks promising, I still have to read the doc properly but here is
> one question for you :)
>
> Are you guys thinking of making this new HR code extensible in way similar
> to the Hook APIs?
>
> We currently use bind10 in production with a custom hook lib that bypasses
> subnet allocation logic and just fetches this info connecting to one of
> our backend systems.
>
> It would be very powerful to make people able to override the source class
> (HostDataSource) or just provide more hook points (I believe this is how
> you call them) so that you could fetch host information from other sources
> that are not mysql databases or any supported source (think about some
> backend system already present in people?s production networks)
>
> I spoke with Vicky a while ago promising that I?ll be working on testing
> my hook code with KEA0.9, unfortunately I didn?t manage to get to that
> point, we are still using bind10, but we should migrate to KEA0.9 starting
> end of October.
>
------------------------------
_______________________________________________
kea-dev mailing list
[email protected]
https://lists.isc.org/mailman/listinfo/kea-dev
End of kea-dev Digest, Vol 7, Issue 1
*************************************
