[aprssig] Delayed packets

AE5PL Lists HamLists at ametx.com
Tue Jun 6 07:21:57 EDT 2006


We have seen this on individual IGates in the past and they have all
been traced to: delays in the IGate serial port driver (usually a USB
serial port is at fault), buffering in the IGate software, replaying
logs in the IGate software, delays in the IGate or server Internet
connection (all IGates going through the server will display this
problem if it is at the server), a rogue server or client with multiple
active connections into APRS-IS, a digipeater delaying packets, etc.

The APRS-IS duplicate checking timer is a 30 second sliding window.
This is enough to eliminate most RF and APRS-IS caused dupes.  It is not
the function of the server dupe check algorithm to try to correct for
malfunctioning client software, malfunctioning digipeaters, or improper
server configurations.  The dupe check algorithm is there to eliminate
legitimate duplicates without damaging things like messaging which
depends on duplicate transmissions to overcome collisions.

Your statement regarding the IGate operator not knowing what could be
causing it is also typical.  Most IGate operators don't know how to use
or don't have network monitoring software on their system.  They also
don't take the time to investigate the various causes and usually say
"its findu's fault".  It doesn't have anything to do with findu.  Findu
is a database population program and doesn't have any affect on APRS-IS
operation.

73,

Pete Loveall AE5PL
mailto:pete at ae5pl.net 

> -----Original Message-----
> From: Robbie, wa9inf
> Posted At: Monday, June 05, 2006 10:51 PM
> Subject: [aprssig] Delayed packets
> 
> There is a station 85 miles from the station I am tracking. 
> Since I am getting the station I am tracking from the APRS-IS 
> via a filter, I see all the I-Gate stations callsigns.. This 
> one station constantly has my tracked station jump backwards 
> with a "delayed" packet by as much as a two minute delay...
> 
> Is this a Findu issue of Dupe checking because too short of 
> dupe checking time cycle, or, how could the I-Gate in 
> question delay I-Gateing by as much as two minutes, and have 
> the tracked station jump backwards with his delayed packet?
> 
> This station does this on a regular basis for any mobile I 
> have tracked via the internet that gets its packet digipeated 
> out to the Rockford area from the Chicagoland area. The sysop 
> said he does not know why his I-Gate could do this.
> 
> I do not know, and I haven't heard about a means of  an 
> I-Gate client being able to dupe check stations it is about 
> to put on the APRS-IS, but I sure can see that in this case, 
> they all should??




More information about the aprssig mailing list