[Xastir] IGating

[James Ewen]

On 8/25/07, Greg Eigsti <greg at eigsti.com> wrote:

> I think we are both agreeing that the DB0ANF tool is better for
> looking at iGate activity than Findu.  Not because the data is
> different but because the DB0ANF tool allow you to view the data
> differently.

Most definitely. Presentation of the data allows you to observe
different aspects of the same information. Findu does not allow much
more than observation of the raw data, or a simple map.

> I am not convinced that the data is *the same*.  Granted it may come
> from the same source but local conditions may cause the data to
> differ slightly (e.g. server problems, connectivity issues, software
> bugs, etc.).

No necessarily local conditions, but quite probably network conditions.

Findu used to run on a single machine, and that machine was the only
collection point for the APRS-IS data. Now there are 3 machines that
findu runs on, with multiple tier-one servers, and a multitude of tier
2 servers. Where exactly DB0ANF pulls his data from is unknown. While
the APRS servers are designed to share all the data, and in an ideal
world no packets would ever get lost, I beleive they all share data
via UDP rather than TCP. This means that there are no acknowledgement
packets sent. This is how APRS works on RF as well, send and forget.
It is possible that a few packets get stomped in the APRS-IS, which is
why I think you see discrepencies between findu and DB0ANF's servers.

> From my experiment(s) yesterday I still am not
> understanding why DB0ANF showed my last packet sent and Findu did not
> - this makes me think there are (or were) problems at Findu.  If you
> can explain away this issue in a way that a dumb monkey like myself
> can understand I'd be grateful!

Hey, I'm just another dumb monkey over here too!

Just stay away from my pile of bananas!

James
VE6SRV