[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[LDM #GUQ-669331]: Both Radar and Satellite feeds stopped downloading abruptly

John,

> I see, thanks.  The slowdown today occurred around 1400Z.

Yes. NEXRAD3 latency went through the roof while IDS|DDPLUS latency was near 
zero.

This is the signature of throttling.

> If that is the case, I should see a ton of TCP Retransmissions and eventual 
> failures in Wireshark during that time.

You wouldn't see that if the throttling is implemented by discarding TCP 
packets. You would simply see regular TCP packets arrive relatively slowly.

> I do see a few "TCP Previous segment not captured" and "TCP Fast 
> Retransmissions" and a ton of "Duplicate ACK"

> Also, ldmping to your server was never interrupted nor slowed in response 
> time.

An ldmping(1) wouldn't be affected because its TCP connection is temporary and 
very low volume.

> Could y'all do a 30 second capture from your similar to this and send it to 
> me to help prove the point to our network folks?

We could do that, but it wouldn't show anything conclusive. From the 
perspective of each endpoint, throttling just looks like a network with lower 
than advertised bandwidth and the TCP layers adjust accordingly.

The best evidence for throttling is the difference between time-series plots of 
latency for of a high-vplume feed (NEXRAD3) and a low-volume feed (IDS|DDPLUS).

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: GUQ-669331
Department: Support LDM
Priority: Normal
Status: Closed
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.