File: //usr/share/webmin/squid/help/edit_delay.html
<header>Delay Pools</header>
<p>
Delay pools provides a means a restrict bandwidth usage of clients. With delay pools, clients can be restricted at the individual level, or in groups. ACLs are used to apply bandwidth limits to users.
<p>
This page displays any existing pools, and their limits. Also, there is a section for global options, which at this time only has one option.
<p>
<b>Initial bucket level</b>
<br>
Because of the way delay pools work, a new client could potentially be permitted to use more than their fair share of bandwidth because their new bucket would be empty. In a congested network, this would mean that other connections from clients with nearly full buckets would be starved of bandwidth for some period of time. This option defaults to 50%, and can likely be left alone, unless your <b>max</b> values are very large (the max value when configuring limits actually applies to the bucket size for the pool).