diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 15:01:30 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 15:01:30 +0000 |
commit | 6beeb1b708550be0d4a53b272283e17e5e35fe17 (patch) | |
tree | 1ce8673d4aaa948e5554000101f46536a1e4cc29 /docs/manual/developer/output-filters.html | |
parent | Initial commit. (diff) | |
download | apache2-6beeb1b708550be0d4a53b272283e17e5e35fe17.tar.xz apache2-6beeb1b708550be0d4a53b272283e17e5e35fe17.zip |
Adding upstream version 2.4.57.upstream/2.4.57upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r-- | docs/manual/developer/output-filters.html | 5 | ||||
-rw-r--r-- | docs/manual/developer/output-filters.html.en | 585 |
2 files changed, 590 insertions, 0 deletions
diff --git a/docs/manual/developer/output-filters.html b/docs/manual/developer/output-filters.html new file mode 100644 index 0000000..ee632a6 --- /dev/null +++ b/docs/manual/developer/output-filters.html @@ -0,0 +1,5 @@ +# GENERATED FROM XML -- DO NOT EDIT + +URI: output-filters.html.en +Content-Language: en +Content-type: text/html; charset=UTF-8 diff --git a/docs/manual/developer/output-filters.html.en b/docs/manual/developer/output-filters.html.en new file mode 100644 index 0000000..cd5cf8c --- /dev/null +++ b/docs/manual/developer/output-filters.html.en @@ -0,0 +1,585 @@ +<?xml version="1.0" encoding="UTF-8"?> +<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> +<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en"><head> +<meta content="text/html; charset=UTF-8" http-equiv="Content-Type" /> +<!-- + XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX + This file is generated from xml source: DO NOT EDIT + XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX + --> +<title>Guide to writing output filters - Apache HTTP Server Version 2.4</title> +<link href="../style/css/manual.css" rel="stylesheet" media="all" type="text/css" title="Main stylesheet" /> +<link href="../style/css/manual-loose-100pc.css" rel="alternate stylesheet" media="all" type="text/css" title="No Sidebar - Default font size" /> +<link href="../style/css/manual-print.css" rel="stylesheet" media="print" type="text/css" /><link rel="stylesheet" type="text/css" href="../style/css/prettify.css" /> +<script src="../style/scripts/prettify.min.js" type="text/javascript"> +</script> + +<link href="../images/favicon.ico" rel="shortcut icon" /></head> +<body id="manual-page"><div id="page-header"> +<p class="menu"><a href="../mod/">Modules</a> | <a href="../mod/directives.html">Directives</a> | <a href="http://wiki.apache.org/httpd/FAQ">FAQ</a> | <a href="../glossary.html">Glossary</a> | <a href="../sitemap.html">Sitemap</a></p> +<p class="apache">Apache HTTP Server Version 2.4</p> +<img alt="" src="../images/feather.png" /></div> +<div class="up"><a href="./"><img title="<-" alt="<-" src="../images/left.gif" /></a></div> +<div id="path"> +<a href="http://www.apache.org/">Apache</a> > <a href="http://httpd.apache.org/">HTTP Server</a> > <a href="http://httpd.apache.org/docs/">Documentation</a> > <a href="../">Version 2.4</a> > <a href="./">Developer Documentation</a></div><div id="page-content"><div id="preamble"><h1>Guide to writing output filters</h1> +<div class="toplang"> +<p><span>Available Languages: </span><a href="../en/developer/output-filters.html" title="English"> en </a></p> +</div> + + <p>There are a number of common pitfalls encountered when writing + output filters; this page aims to document best practice for + authors of new or existing filters.</p> + + <p>This document is applicable to both version 2.0 and version 2.2 + of the Apache HTTP Server; it specifically targets + <code>RESOURCE</code>-level or <code>CONTENT_SET</code>-level + filters though some advice is generic to all types of filter.</p> + </div> +<div id="quickview"><a href="https://www.apache.org/foundation/contributing.html" class="badge"><img src="https://www.apache.org/images/SupportApache-small.png" alt="Support Apache!" /></a><ul id="toc"><li><img alt="" src="../images/down.gif" /> <a href="#basics">Filters and bucket brigades</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#invocation">Filter invocation</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#brigade">Brigade structure</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#buckets">Processing buckets</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#filtering">Filtering brigades</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#state">Maintaining state</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#buffer">Buffering buckets</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#nonblock">Non-blocking bucket reads</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#rules">Ten rules for output filters</a></li> +<li><img alt="" src="../images/down.gif" /> <a href="#usecase1">Use case: buffering in mod_ratelimit</a></li> +</ul><h3>See also</h3><ul class="seealso"><li><a href="#comments_section">Comments</a></li></ul></div> +<div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="basics" id="basics">Filters and bucket brigades</a></h2> + + + <p>Each time a filter is invoked, it is passed a <em>bucket + brigade</em>, containing a sequence of <em>buckets</em> which + represent both data content and metadata. Every bucket has a + <em>bucket type</em>; a number of bucket types are defined and + used by the <code>httpd</code> core modules (and the + <code>apr-util</code> library which provides the bucket brigade + interface), but modules are free to define their own types.</p> + + <div class="note">Output filters must be prepared to process + buckets of non-standard types; with a few exceptions, a filter + need not care about the types of buckets being filtered.</div> + + <p>A filter can tell whether a bucket represents either data or + metadata using the <code>APR_BUCKET_IS_METADATA</code> macro. + Generally, all metadata buckets should be passed down the filter + chain by an output filter. Filters may transform, delete, and + insert data buckets as appropriate.</p> + + <p>There are two metadata bucket types which all filters must pay + attention to: the <code>EOS</code> bucket type, and the + <code>FLUSH</code> bucket type. An <code>EOS</code> bucket + indicates that the end of the response has been reached and no + further buckets need be processed. A <code>FLUSH</code> bucket + indicates that the filter should flush any buffered buckets (if + applicable) down the filter chain immediately.</p> + + <div class="note"><code>FLUSH</code> buckets are sent when the + content generator (or an upstream filter) knows that there may be + a delay before more content can be sent. By passing + <code>FLUSH</code> buckets down the filter chain immediately, + filters ensure that the client is not kept waiting for pending + data longer than necessary.</div> + + <p>Filters can create <code>FLUSH</code> buckets and pass these + down the filter chain if desired. Generating <code>FLUSH</code> + buckets unnecessarily, or too frequently, can harm network + utilisation since it may force large numbers of small packets to + be sent, rather than a small number of larger packets. The + section on <a href="#nonblock">Non-blocking bucket reads</a> + covers a case where filters are encouraged to generate + <code>FLUSH</code> buckets.</p> + + <div class="example"><h3>Example bucket brigade</h3><p><code> + HEAP FLUSH FILE EOS</code></p></div> + + <p>This shows a bucket brigade which may be passed to a filter; it + contains two metadata buckets (<code>FLUSH</code> and + <code>EOS</code>), and two data buckets (<code>HEAP</code> and + <code>FILE</code>).</p> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="invocation" id="invocation">Filter invocation</a></h2> + + + <p>For any given request, an output filter might be invoked only + once and be given a single brigade representing the entire response. + It is also possible that the number of times a filter is invoked + for a single response is proportional to the size of the content + being filtered, with the filter being passed a brigade containing + a single bucket each time. Filters must operate correctly in + either case.</p> + + <div class="warning">An output filter which allocates long-lived + memory every time it is invoked may consume memory proportional to + response size. Output filters which need to allocate memory + should do so once per response; see <a href="#state">Maintaining + state</a> below.</div> + + <p>An output filter can distinguish the final invocation for a + given response by the presence of an <code>EOS</code> bucket in + the brigade. Any buckets in the brigade after an EOS should be + ignored.</p> + + <p>An output filter should never pass an empty brigade down the + filter chain. To be defensive, filters should be prepared to + accept an empty brigade, and should return success without passing + this brigade on down the filter chain. The handling of an empty + brigade should have no side effects (such as changing any state + private to the filter).</p> + + <div class="example"><h3>How to handle an empty brigade</h3><pre class="prettyprint lang-c">apr_status_t dummy_filter(ap_filter_t *f, apr_bucket_brigade *bb) +{ + if (APR_BRIGADE_EMPTY(bb)) { + return APR_SUCCESS; + } + ...</pre> +</div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="brigade" id="brigade">Brigade structure</a></h2> + + + <p>A bucket brigade is a doubly-linked list of buckets. The list + is terminated (at both ends) by a <em>sentinel</em> which can be + distinguished from a normal bucket by comparing it with the + pointer returned by <code>APR_BRIGADE_SENTINEL</code>. The list + sentinel is in fact not a valid bucket structure; any attempt to + call normal bucket functions (such as + <code>apr_bucket_read</code>) on the sentinel will have undefined + behaviour (i.e. will crash the process).</p> + + <p>There are a variety of functions and macros for traversing and + manipulating bucket brigades; see the <a href="http://apr.apache.org/docs/apr-util/trunk/group___a_p_r___util___bucket___brigades.html">apr_buckets.h</a> + header for complete coverage. Commonly used macros include:</p> + + <dl> + <dt><code>APR_BRIGADE_FIRST(bb)</code></dt> + <dd>returns the first bucket in brigade bb</dd> + + <dt><code>APR_BRIGADE_LAST(bb)</code></dt> + <dd>returns the last bucket in brigade bb</dd> + + <dt><code>APR_BUCKET_NEXT(e)</code></dt> + <dd>gives the next bucket after bucket e</dd> + + <dt><code>APR_BUCKET_PREV(e)</code></dt> + <dd>gives the bucket before bucket e</dd> + + </dl> + + <p>The <code>apr_bucket_brigade</code> structure itself is + allocated out of a pool, so if a filter creates a new brigade, it + must ensure that memory use is correctly bounded. A filter which + allocates a new brigade out of the request pool + (<code>r->pool</code>) on every invocation, for example, will fall + foul of the <a href="#invocation">warning above</a> concerning + memory use. Such a filter should instead create a brigade on the + first invocation per request, and store that brigade in its <a href="#state">state structure</a>.</p> + + <div class="warning"><p>It is generally never advisable to use + <code>apr_brigade_destroy</code> to "destroy" a brigade unless + you know for certain that the brigade will never be used + again, even then, it should be used rarely. The + memory used by the brigade structure will not be released by + calling this function (since it comes from a pool), but the + associated pool cleanup is unregistered. Using + <code>apr_brigade_destroy</code> can in fact cause memory leaks; + if a "destroyed" brigade contains buckets when its + containing pool is destroyed, those buckets will <em>not</em> be + immediately destroyed.</p> + + <p>In general, filters should use <code>apr_brigade_cleanup</code> + in preference to <code>apr_brigade_destroy</code>.</p></div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="buckets" id="buckets">Processing buckets</a></h2> + + + + <p>When dealing with non-metadata buckets, it is important to + understand that the "<code>apr_bucket *</code>" object is an + abstract <em>representation</em> of data:</p> + + <ol> + <li>The amount of data represented by the bucket may or may not + have a determinate length; for a bucket which represents data of + indeterminate length, the <code>->length</code> field is set to + the value <code>(apr_size_t)-1</code>. For example, buckets of + the <code>PIPE</code> bucket type have an indeterminate length; + they represent the output from a pipe.</li> + + <li>The data represented by a bucket may or may not be mapped + into memory. The <code>FILE</code> bucket type, for example, + represents data stored in a file on disk.</li> + </ol> + + <p>Filters read the data from a bucket using the + <code>apr_bucket_read</code> function. When this function is + invoked, the bucket may <em>morph</em> into a different bucket + type, and may also insert a new bucket into the bucket brigade. + This must happen for buckets which represent data not mapped into + memory.</p> + + <p>To give an example; consider a bucket brigade containing a + single <code>FILE</code> bucket representing an entire file, 24 + kilobytes in size:</p> + + <div class="example"><p><code>FILE(0K-24K)</code></p></div> + + <p>When this bucket is read, it will read a block of data from the + file, morph into a <code>HEAP</code> bucket to represent that + data, and return the data to the caller. It also inserts a new + <code>FILE</code> bucket representing the remainder of the file; + after the <code>apr_bucket_read</code> call, the brigade looks + like:</p> + + <div class="example"><p><code>HEAP(8K) FILE(8K-24K)</code></p></div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="filtering" id="filtering">Filtering brigades</a></h2> + + + <p>The basic function of any output filter will be to iterate + through the passed-in brigade and transform (or simply examine) + the content in some manner. The implementation of the iteration + loop is critical to producing a well-behaved output filter.</p> + + <p>Taking an example which loops through the entire brigade as + follows:</p> + + <div class="example"><h3>Bad output filter -- do not imitate!</h3><pre class="prettyprint lang-c">apr_bucket *e = APR_BRIGADE_FIRST(bb); +const char *data; +apr_size_t length; + +while (e != APR_BRIGADE_SENTINEL(bb)) { + apr_bucket_read(e, &data, &length, APR_BLOCK_READ); + e = APR_BUCKET_NEXT(e); +} + +return ap_pass_brigade(bb);</pre> +</div> + + <p>The above implementation would consume memory proportional to + content size. If passed a <code>FILE</code> bucket, for example, + the entire file contents would be read into memory as each + <code>apr_bucket_read</code> call morphed a <code>FILE</code> + bucket into a <code>HEAP</code> bucket.</p> + + <p>In contrast, the implementation below will consume a fixed + amount of memory to filter any brigade; a temporary brigade is + needed and must be allocated only once per response, see the <a href="#state">Maintaining state</a> section.</p> + + <div class="example"><h3>Better output filter</h3><pre class="prettyprint lang-c">apr_bucket *e; +const char *data; +apr_size_t length; + +while ((e = APR_BRIGADE_FIRST(bb)) != APR_BRIGADE_SENTINEL(bb)) { + rv = apr_bucket_read(e, &data, &length, APR_BLOCK_READ); + if (rv) ...; + /* Remove bucket e from bb. */ + APR_BUCKET_REMOVE(e); + /* Insert it into temporary brigade. */ + APR_BRIGADE_INSERT_HEAD(tmpbb, e); + /* Pass brigade downstream. */ + rv = ap_pass_brigade(f->next, tmpbb); + if (rv) ...; + apr_brigade_cleanup(tmpbb); +}</pre> +</div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="state" id="state">Maintaining state</a></h2> + + + + <p>A filter which needs to maintain state over multiple + invocations per response can use the <code>->ctx</code> field of + its <code>ap_filter_t</code> structure. It is typical to store a + temporary brigade in such a structure, to avoid having to allocate + a new brigade per invocation as described in the <a href="#brigade">Brigade structure</a> section.</p> + + <div class="example"><h3>Example code to maintain filter state</h3><pre class="prettyprint lang-c">struct dummy_state { + apr_bucket_brigade *tmpbb; + int filter_state; + ... +}; + +apr_status_t dummy_filter(ap_filter_t *f, apr_bucket_brigade *bb) +{ + struct dummy_state *state; + + state = f->ctx; + if (state == NULL) { + + /* First invocation for this response: initialise state structure. + */ + f->ctx = state = apr_palloc(f->r->pool, sizeof *state); + + state->tmpbb = apr_brigade_create(f->r->pool, f->c->bucket_alloc); + state->filter_state = ...; + } + ...</pre> +</div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="buffer" id="buffer">Buffering buckets</a></h2> + + + <p>If a filter decides to store buckets beyond the duration of a + single filter function invocation (for example storing them in its + <code>->ctx</code> state structure), those buckets must be <em>set + aside</em>. This is necessary because some bucket types provide + buckets which represent temporary resources (such as stack memory) + which will fall out of scope as soon as the filter chain completes + processing the brigade.</p> + + <p>To setaside a bucket, the <code>apr_bucket_setaside</code> + function can be called. Not all bucket types can be setaside, but + if successful, the bucket will have morphed to ensure it has a + lifetime at least as long as the pool given as an argument to the + <code>apr_bucket_setaside</code> function.</p> + + <p>Alternatively, the <code>ap_save_brigade</code> function can be + used, which will move all the buckets into a separate brigade + containing buckets with a lifetime as long as the given pool + argument. This function must be used with care, taking into + account the following points:</p> + + <ol> + <li>On return, <code>ap_save_brigade</code> guarantees that all + the buckets in the returned brigade will represent data mapped + into memory. If given an input brigade containing, for example, + a <code>PIPE</code> bucket, <code>ap_save_brigade</code> will + consume an arbitrary amount of memory to store the entire output + of the pipe.</li> + + <li>When <code>ap_save_brigade</code> reads from buckets which + cannot be setaside, it will always perform blocking reads, + removing the opportunity to use <a href="#nonblock">Non-blocking + bucket reads</a>.</li> + + <li>If <code>ap_save_brigade</code> is used without passing a + non-NULL "<code>saveto</code>" (destination) brigade parameter, + the function will create a new brigade, which may cause memory + use to be proportional to content size as described in the <a href="#brigade">Brigade structure</a> section.</li> + </ol> + + <div class="warning">Filters must ensure that any buffered data is + processed and passed down the filter chain during the last + invocation for a given response (a brigade containing an EOS + bucket). Otherwise such data will be lost.</div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="nonblock" id="nonblock">Non-blocking bucket reads</a></h2> + + + <p>The <code>apr_bucket_read</code> function takes an + <code>apr_read_type_e</code> argument which determines whether a + <em>blocking</em> or <em>non-blocking</em> read will be performed + from the data source. A good filter will first attempt to read + from every data bucket using a non-blocking read; if that fails + with <code>APR_EAGAIN</code>, then send a <code>FLUSH</code> + bucket down the filter chain, and retry using a blocking read.</p> + + <p>This mode of operation ensures that any filters further down the + filter chain will flush any buffered buckets if a slow content + source is being used.</p> + + <p>A CGI script is an example of a slow content source which is + implemented as a bucket type. <code class="module"><a href="../mod/mod_cgi.html">mod_cgi</a></code> will send + <code>PIPE</code> buckets which represent the output from a CGI + script; reading from such a bucket will block when waiting for the + CGI script to produce more output.</p> + + <div class="example"><h3>Example code using non-blocking bucket reads</h3><pre class="prettyprint lang-c">apr_bucket *e; +apr_read_type_e mode = APR_NONBLOCK_READ; + +while ((e = APR_BRIGADE_FIRST(bb)) != APR_BRIGADE_SENTINEL(bb)) { + apr_status_t rv; + + rv = apr_bucket_read(e, &data, &length, mode); + if (rv == APR_EAGAIN && mode == APR_NONBLOCK_READ) { + + /* Pass down a brigade containing a flush bucket: */ + APR_BRIGADE_INSERT_TAIL(tmpbb, apr_bucket_flush_create(...)); + rv = ap_pass_brigade(f->next, tmpbb); + apr_brigade_cleanup(tmpbb); + if (rv != APR_SUCCESS) return rv; + + /* Retry, using a blocking read. */ + mode = APR_BLOCK_READ; + continue; + } + else if (rv != APR_SUCCESS) { + /* handle errors */ + } + + /* Next time, try a non-blocking read first. */ + mode = APR_NONBLOCK_READ; + ... +}</pre> +</div> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="rules" id="rules">Ten rules for output filters</a></h2> + + + <p>In summary, here is a set of rules for all output filters to + follow:</p> + + <ol> + <li>Output filters should not pass empty brigades down the filter + chain, but should be tolerant of being passed empty + brigades.</li> + + <li>Output filters must pass all metadata buckets down the filter + chain; <code>FLUSH</code> buckets should be respected by passing + any pending or buffered buckets down the filter chain.</li> + + <li>Output filters should ignore any buckets following an + <code>EOS</code> bucket.</li> + + <li>Output filters must process a fixed amount of data at a + time, to ensure that memory consumption is not proportional to + the size of the content being filtered.</li> + + <li>Output filters should be agnostic with respect to bucket + types, and must be able to process buckets of unfamiliar + type.</li> + + <li>After calling <code>ap_pass_brigade</code> to pass a brigade + down the filter chain, output filters should call + <code>apr_brigade_cleanup</code> to ensure the brigade is empty + before reusing that brigade structure; output filters should + never use <code>apr_brigade_destroy</code> to "destroy" + brigades.</li> + + <li>Output filters must <em>setaside</em> any buckets which are + preserved beyond the duration of the filter function.</li> + + <li>Output filters must not ignore the return value of + <code>ap_pass_brigade</code>, and must return appropriate errors + back up the filter chain.</li> + + <li>Output filters must only create a fixed number of bucket + brigades for each response, rather than one per invocation.</li> + + <li>Output filters should first attempt non-blocking reads from + each data bucket, and send a <code>FLUSH</code> bucket down the + filter chain if the read blocks, before retrying with a blocking + read.</li> + + </ol> + + </div><div class="top"><a href="#page-header"><img alt="top" src="../images/up.gif" /></a></div> +<div class="section"> +<h2><a name="usecase1" id="usecase1">Use case: buffering in mod_ratelimit</a></h2> + + <p>The <a href="http://svn.apache.org/r1833875">r1833875</a> change is a good + example to show what buffering and keeping state means in the context of an + output filter. In this use case, a user asked on the users' mailing list a + interesting question about why <code class="module"><a href="../mod/mod_ratelimit.html">mod_ratelimit</a></code> seemed not to + honor its setting with proxied content (either rate limiting at a different + speed or simply not doing it at all). Before diving deep into the solution, + it is better to explain on a high level how <code class="module"><a href="../mod/mod_ratelimit.html">mod_ratelimit</a></code> works. + The trick is really simple: take the rate limit settings and calculate a + chunk size of data to flush every 200ms to the client. For example, let's imagine + that to set <code>rate-limit 60</code> in our config, these are the high level + steps to find the chunk size:</p> + <pre class="prettyprint lang-c">/* milliseconds to wait between each flush of data */ +RATE_INTERVAL_MS = 200; +/* rate limit speed in b/s */ +speed = 60 * 1024; +/* final chunk size is 12228 bytes */ +chunk_size = (speed / (1000 / RATE_INTERVAL_MS));</pre> + + <p>If we apply this calculation to a bucket brigade carrying 38400 bytes, it means + that the filter will try to do the following:</p> + <ol> + <li>Split the 38400 bytes in chunks of maximum 12228 bytes each.</li> + <li>Flush the first 12228 chunk of bytes and sleep 200ms.</li> + <li>Flush the second 12228 chunk of bytes and sleep 200ms.</li> + <li>Flush the third 12228 chunk of bytes and sleep 200ms.</li> + <li>Flush the remaining 1716 bytes.</li> + </ol> + <p>The above pseudo code works fine if the output filter handles only one brigade + for each response, but it might happen that it needs to be called multiple times + with different brigade sizes as well. The former use case is for example when + httpd directly serves some content, like a static file: the bucket brigade + abstraction takes care of handling the whole content, and rate limiting + works nicely. But if the same static content is served via mod_proxy_http (for + example a backend is serving it rather than httpd) then the content generator + (in this case mod_proxy_http) may use a maximum buffer size and then send data + as bucket brigades to the output filters chain regularly, triggering of course + multiple calls to <code class="module"><a href="../mod/mod_ratelimit.html">mod_ratelimit</a></code>. If the reader tries to execute the pseudo code + assuming multiple calls to the output filter, each one requiring to process + a bucket brigade of 38400 bytes, then it is easy to spot some + anomalies:</p> + <ol> + <li>Between the last flush of a brigade and the first one of the next, + there is no sleep.</li> + <li>Even if the sleep was forced after the last flush, then that chunk size + would not be the ideal size (1716 bytes instead of 12228) and the final client's speed + would quickly become different than what set in the httpd's config.</li> + </ol> + <p>In this case, two things might help:</p> + <ol> + <li>Use the ctx internal data structure, initialized by <code class="module"><a href="../mod/mod_ratelimit.html">mod_ratelimit</a></code> + for each response handling cycle, to "remember" when the last sleep was + performed across multiple invocations, and act accordingly.</li> + <li>If a bucket brigade is not splittable into a finite number of chunk_size + blocks, store the remaining bytes (located in the tail of the bucket brigade) + in a temporary holding area (namely another bucket brigade) and then use + <code>ap_save_brigade</code> to set them aside. + These bytes will be prepended to the next bucket brigade that will be handled + in the subsequent invocation.</li> + <li>Avoid the previous logic if the bucket brigade that is currently being + processed contains the end of stream bucket (EOS). There is no need to sleep + or buffering data if the end of stream is reached.</li> + </ol> + <p>The commit linked in the beginning of the section contains also a bit of code + refactoring so it is not trivial to read during the first pass, but the overall + idea is basically what written up to now. The goal of this section is not to + cause a headache to the reader trying to read C code, but to put him/her into + the right mindset needed to use efficiently the tools offered by the httpd's + filter chain toolset.</p> + </div></div> +<div class="bottomlang"> +<p><span>Available Languages: </span><a href="../en/developer/output-filters.html" title="English"> en </a></p> +</div><div class="top"><a href="#page-header"><img src="../images/up.gif" alt="top" /></a></div><div class="section"><h2><a id="comments_section" name="comments_section">Comments</a></h2><div class="warning"><strong>Notice:</strong><br />This is not a Q&A section. Comments placed here should be pointed towards suggestions on improving the documentation or server, and may be removed by our moderators if they are either implemented or considered invalid/off-topic. Questions on how to manage the Apache HTTP Server should be directed at either our IRC channel, #httpd, on Libera.chat, or sent to our <a href="https://httpd.apache.org/lists.html">mailing lists</a>.</div> +<script type="text/javascript"><!--//--><![CDATA[//><!-- +var comments_shortname = 'httpd'; +var comments_identifier = 'http://httpd.apache.org/docs/2.4/developer/output-filters.html'; +(function(w, d) { + if (w.location.hostname.toLowerCase() == "httpd.apache.org") { + d.write('<div id="comments_thread"><\/div>'); + var s = d.createElement('script'); + s.type = 'text/javascript'; + s.async = true; + s.src = 'https://comments.apache.org/show_comments.lua?site=' + comments_shortname + '&page=' + comments_identifier; + (d.getElementsByTagName('head')[0] || d.getElementsByTagName('body')[0]).appendChild(s); + } + else { + d.write('<div id="comments_thread">Comments are disabled for this page at the moment.<\/div>'); + } +})(window, document); +//--><!]]></script></div><div id="footer"> +<p class="apache">Copyright 2023 The Apache Software Foundation.<br />Licensed under the <a href="http://www.apache.org/licenses/LICENSE-2.0">Apache License, Version 2.0</a>.</p> +<p class="menu"><a href="../mod/">Modules</a> | <a href="../mod/directives.html">Directives</a> | <a href="http://wiki.apache.org/httpd/FAQ">FAQ</a> | <a href="../glossary.html">Glossary</a> | <a href="../sitemap.html">Sitemap</a></p></div><script type="text/javascript"><!--//--><![CDATA[//><!-- +if (typeof(prettyPrint) !== 'undefined') { + prettyPrint(); +} +//--><!]]></script> +</body></html>
\ No newline at end of file |