[hydra-dev] Still confused, but at a higher level

David Ruggles thedavidfactor at gmail.com
Mon Apr 12 09:22:29 CDT 2010


After sitting through the meeting and absorbing everything I could;
here's how I see the breakdown:

Asterisk
The best PBX money can (or can't) buy:
* (relatively) easy to use: mostly text based config files, many PBX
applications. (i.e. creating a full featured conference or voicemail
solution takes editing two, maybe three, text files out of the box)

* structured: leads to the ease of use, but also causes arbitrary
limits: It creates a situation where you spend a lot of time trying to
code around limits without breaking any other pre-existing
functionality.

* (relatively) limited scalability: I heard the 10,000 call limit
thrown around, but I got the feeling that may have been created in a
lab, in my (limited) real-world experience where my company is pushing
8 - 10 thousand concurrent calls with G.729, G.711 and some
transcoding we limit concurrent calls on a single Asterisk instance to
300 - 500 so quality is impacted. However we can run more then a
single instance of Asterisk on the same box so our per box call limit
is in the 1200 - 2000 range.


Hydra
The best Telecommunications[1] Layer 7 router/gateway money can (or can't) buy:
* not a PBX: I don't see it trying to incorporate the conferencing
features, the voicemail features or the other more traditional PBX
type features, at least not in an as user friendly way as Asterisk.
For example a conference service provider might use Hyra to be able to
support 100s of thousands of users in 10s of thousands of conferences.
But if a user is just trying to add conferencing for their office they
would use Asterisk

* Scalable: I feel this is Hydra's defining attribute. Whatever box it
ends up coming in, it should be scalable to the moon out of the box
simply by adding hardware. And yes the hardware may be virtual. So
what Cloud computing is doing to the data-center, Hyra will do to the
telecom-center.

* Partnering: the ability to loosely couple and federate disparate
telecommunications systems provides new methods for customers and
providers to interact in a seamless manner


This is a rough outline of what I heard, saw and took away. I also
heard John Todd (I believe) talk about the "long tail" of Asterisk. I
see Hydra as being concentrated at the end of that long tail, with a
long tail of it's own that points back at Asterisk. So yes, there will
be cross over, there will be places where it might be hard to point to
one or the other and definitively say which one is "best" for a
particular purpose, but with a broad bush I think the points I've laid
out above do provide a significant amount of distinction.

Now I submit this to be torn apart by everyone :-D



[1] I know the word Telecommunications seems to have baggage, but I
use it in the form defined by wikipedia: "the transmission of messages
over significant distances for the purpose of communication."




More information about the asterisk-scf-dev mailing list