squid限速

squid有個delay_pool，可以做限速，雖然效果不太準~（嗯，就像限制并發連接配接數的maxconn一樣） 首先搬個老虎皮做大旗——《Squid: The Definitive Guide》的相關段落：

The buckets don’t actually store bandwidth (e.g., 100 Kbit/s), but 

rather some amount of traffic (e.g., 384 KB). Squid adds some 

amount of traffic to the buckets each second. Cache clients take 

some amount of traffic out when they receive data from an upstream 

source (origin server or neighbor). 

The size of a bucket determines how much burst bandwidth is 

available to a client. If a bucket starts out full, a client can 

take as much traffic as it needs until the bucket becomes empty. 

The client then receives traffic allotments at the fill rate. 

The mapping between Squid clients and actual buckets is a bit 

complicated. Squid uses three different constructs to do it: access 

rules, delay pool classes, and types of buckets. First, Squid 

checks a client request against the delay_access list. If the 

request is a match, it points to a particular delay pool. Each 

delay pool has a class: 1, 2, or 3. The classes determine which 

types of buckets are in use. Squid has three types of buckets: 

aggregate, individual, and network: 

A class 1 pool has a single aggregate bucket. 

A class 2 pool has an aggregate bucket and 256 individual 

buckets. 

A class 3 pool has an aggregate bucket, 256 network buckets, and 

65,536 individual buckets. 

As you can probably guess, the individual and network buckets 

correspond to IP address octets. In a class 2 pool, the individual 

bucket is determined by the last octet of the client’s IPv4 

address. In a class 3 pool, the network bucket is determined by the 

third octet, and the individual bucket by the third and fourth 

octets. 

For the class 2 and 3 delay pools, you can disable buckets you 

don’t want to use. For example, you can define a class 2 pool with 

only individual buckets by disabling the aggregate bucket. 

When a request goes through a pool with more than one bucket type, 

it takes bandwidth from all buckets. For example, consider a class 

3 pool with aggregate, network, and individual buckets. If the 

individual bucket has 20 KB, the network bucket 30 KB, but the 

aggregate bucket only 2 KB, the client receives only a 2-KB 

allotment. Even though some buckets have plenty of traffic, the 

client is limited by the bucket with the smallest amount. 

C.2 Configuring Squid 

Before you can use delay pools, you must enable the feature when 

compiling. Use the —enable-delay-pools option when running 

./configure. You can then use the following directives to set up 

the delay pools. 

C.2.1 delay_pools 

The delay_pools directive tells Squid how many pools you want to 

define. It should go before any other delay pool-configuration 

directives in squid.conf. For example, if you want to have five 

delay pools: 

delay_pools 5 

The next two directives actually define each pool’s class and other 

characteristics. 

C.2.2 delay_class 

You must use this directive to define the class for each pool. For 

example, if the first pool is class 3: 

delay_class 1 3 

Similarly, if the fourth pool is class 2: 

delay_class 4 2 

In theory, you should have one delay_class line for each pool. 

However, if you skip or omit a particular pool, Squid doesn’t 

complain. 

C.2.3 delay_parameters 

Finally, this is where you define the interesting delay pool 

parameters. For each pool, you must tell Squid the fill rate and 

maximum size for each type of bucket. The syntax is: 

delay_parameters N rate/size [rate/size [rate/size]] 

The rate value is given in bytes per second, and size in total 

bytes. If you think of rate in terms of bits per second, you must 

remember to divide by 8. 

Note that if you divide the size by the rate, you’ll know how long 

it takes (number of seconds) the bucket to go from empty to full 

when there are no clients using it. 

A class 1 pool has just one bucket and might look like this: 

delay_class 2 1 

delay_parameters 2 2000/8000 

For a class 2 pool, the first bucket is the aggregate, and the 

second is the group of individual buckets. For example: 

delay_parameters 4 7000/15000 3000/4000 

Similarly, for a class 3 pool, the aggregate bucket is first, the 

network buckets are second, and the individual buckets are 

third: 

delay_parameters 1 7000/15000 3000/4000 1000/2000 

C.2.4 delay_initial_bucket_level 

This directive sets the initial level for all buckets when Squid 

first starts or is reconfigured. It also applies to individual and 

network 

buckets, which aren’t created until first referenced. The value is 

a percentage. For example: 

delay_initial_bucket_level 75% 

In this case, each newly created bucket is initially filled to 75% 

of its maximum size. 

C.2.5 delay_access 

This list of access rules determines which requests go through 

which delay pools. Requests that are allowed go through the delay 

pools, while those that are denied aren’t delayed at all. If you 

don’t have any delay_access rules, Squid doesn’t delay any 

requests. 

The syntax for delay_access is similar to the other access rule 

lists (see Section 6.2), except that you must put a pool number 

before the allow or deny keyword. For example: 

delay_access 1 allow TheseUsers 

delay_access 2 allow OtherUsers 

Internally, Squid stores a separate access rule list for each delay 

pool. If a request is allowed by a pool’s rules, Squid uses that 

pool and stops searching. If a request is denied, however, Squid 

continues examining the rules for remaining pools. In other words, 

a deny rule causes Squid to stop searching the rules for a single 

pool but not for all pools. 

C.2.6 cache_peer no-delay Option 

The cache_peer directive has a no-delay option. If set, it makes 

Squid bypass the delay pools for any requests sent to that 

neighbor.

然後說老實話：我也看不太懂…… 隻好貼一些百度出來的結果： class類型1為單個IP位址流量 

class類型2為C類網段中的每個IP位址流量 

class類型3為B類網段中的每個C類網段中的每個IP位址流量 

具體的說： 類型1隻有一個總帶寬流量實際也就是這個IP位址的流量 

delay_parameters 1 64000/64000 

類型2有兩個帶寬流量參數，第一個為整個C類型網段流量，第二個為每個IP流量 

delay_parameters 1 -1/-1 64000/64000 

類型3有三個帶寬流量參數,第一個為整個B類網總流量，第二個為每個B類網段中的C類網段總流量,第三個為了B類網段中每個C類網段中的每個IP流量 

delay_parameters 1 -1/-1 -1/-1 64000/64000 

但似乎我還沒百度到誰用class為2或者3的。一般大家都隻用1…… 舉個例子： 兩個域名，分别限制網民下載下傳速度為50kb/s和100kb/s。配置如下：

據網友的測試，當限速配置為20000/20000即20000/1024=19.53kb/s的時候，實際的下載下傳速度大概在11-15kb/s之間。