Re: Problems with some web sites (tuning?)

[Carl Johnson <carlj@peak.org>]

"Barclay, Daniel" <daniel@fgm.com> writes:

> Carl Johnson wrote:
>> owens@netptc.net writes:
>> 
>>> ... The
>>> timeout or "lockup" can indicate that the packets cannot be
>>> reassembled at the destination (your computer) and the TCP protocol
>>> times out waiting for one or more missing packets.
>> 
>> That makes sense to me, but why is it only a very few web sites?  I
>> haven't heard complaints about poor wikipedia access, so it appears
>> that most other people don't have problems with that either.  One idea
>> I thought of is that maybe they have very tight timeout limits, and
>> since I am on dialup, I often exceed those limits and they then drop
>> packets.  
>
> That theory seems reasonable--Google used to exhibit similar behavior.
>
> When I was using PPP, if I posted a Google search when nothing was
> going on on my network connection, it would work fine.  However, if I
> submitted a search when some other connection was sending packets,
> causing the packet to and from Google to take a couple of seconds, the
> Google connection would time out.  Presumably they were trying to defend
> against SYN flood attacks.  I wrote to Google and a couple of months
> later the problem stopped happening.
>
>
>> Any idea how I could trace something like dropped packets?  
>
> It might not be _dropped_ packets--it might be high packet latency.

That's more or less what I was thinking.  High latency in my dialup
line delays the packets enough that wikipedia has already timed out
the connection, so it ignores the replies.  So the solution would be
to use wikipedia when my dialup line is otherwise unloaded.

Thanks for the information.
-- 
Carl Johnson		carlj@peak.org