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

[LDM #XXW-990118]: LDM - Not your average latency problem

Mike,

[Regarding the file /etc/hosts]
> OK, this has been done. In addition, I have increased the total feed by
> adding back in the NIMAGE feed.
> 
[Regarding the hypothesis that LDM performance is affected by non-LDM traffic]
> I see your point, but this only makes sense if the pipe is limited between
> STI and us....because the only other traffic on their Internet connection
> is some minor VPN traffic. STI's regular Internet traffic is on a
> completely different pipe.
> 
> I have attached two MRTG graphs from STI's pipe, from last week, which show
> how the pipe is full and then not (note: they increased the pipe from ~360
> to 500 Kbps last Wednesday to try to remedy the situation, but this did not
> help) . But the problem is, when the pipe is not full we are still seeing
> large latencies and some of the feeds just stop. This does not seem
> consistent with your explanation, unless of course I'm missing something.

The MRTG graphs do show a strong diurnal signal, so I agree that LDM 
performance should be good during local nighttime (which it isn't).  We'll 
investigate further.

In the MRTG plots, "incoming" and "outgoing" refer to what, exactly?  In 
similar plots here, the sense is the reverse of what one might naively expect.

In the meantime, please send us the most recent MRTG graph so that we can 
correlate it with the current
rtstats(1) plots.

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: XXW-990118
Department: Support LDM
Priority: Normal
Status: On Hold