Re: bandwidth spikes

From: Henrik Nordstrom <>
Date: Sun, 31 Jan 1999 18:57:28 +0100

Ross Wheeler wrote:

> simply not viable. Thus, when a dial-up customer at say, 28K8, requests a
> large file through squid, and especially if that file is at a well
> connected site (say, in the proxy at an upstream provider), squid will
> suck the file down ASAP.

Not true.

It will suck down 2*TCP_WINDOW+64KB of data, then things slows down to
the client speed. If delay pools are used then this is further
restricted to 1*TCP_WINDOW + delay_pool.

However, if you havent enabled quick_abort (quick_abort_min 0) in
squid.conf, then Squid continues to download the object if the client
aborts the connection, and this is done at full speed.

Bandwidth spikes may also be caused by Range requests in Squid 2.0. This
has been addressed in Squid 2.1 by the squid.conf range_offset_limit
directive. Hopefully the problem with storage of partial objects is
addressed soon so that no bandwidth spikes can be caused by Range

Henrik Nordstrom
Spare time Squid hacker
Received on Sun Jan 31 1999 - 11:21:27 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:44:11 MST