diff options
Diffstat (limited to 'erts/doc/src/counters.xml')
| -rw-r--r-- | erts/doc/src/counters.xml | 170 |
1 files changed, 170 insertions, 0 deletions
diff --git a/erts/doc/src/counters.xml b/erts/doc/src/counters.xml new file mode 100644 index 0000000000..ba4a22759f --- /dev/null +++ b/erts/doc/src/counters.xml @@ -0,0 +1,170 @@ +<?xml version="1.0" encoding="utf-8" ?> +<!DOCTYPE erlref SYSTEM "erlref.dtd"> + +<erlref> + <header> + <copyright> + <year>2018</year> + <holder>Ericsson AB. All Rights Reserved.</holder> + </copyright> + <legalnotice> + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. + </legalnotice> + + <title>counters</title> + </header> + <module>counters</module> + <modulesummary>Counter Functions</modulesummary> + <description> + <p>This module provides a set of functions to do operations towards + shared mutable counter variables. The implementation does not utilize any + software level locking, which makes it very efficient for concurrent + access. The counters are organized into arrays with the follwing + semantics:</p> + <list type="bulleted"> + <item> + <p>Counters are 64 bit signed integers.</p> + </item> + <item> + <p>Counters wrap around at overflow and underflow operations.</p> + </item> + <item><p>Counters are initialized to zero and can then only be written to + by adding or subtracting.</p> + </item> + <item> + <p>Write operations guarantee atomicity. No intermediate results can be + seen from a single write operation.</p> + </item> + <item> + <p>Two types of counter arrays can be created with options <c>atomics</c> or + <c>write_concurrency</c>. The <c>atomics</c> counters have good allround + performance with nice consistent semantics while + <c>write_concurrency</c> counters offers even better concurrent + write performance at the expense of some potential read + inconsistencies. See <seealso marker="#new/2"><c>new/2</c></seealso>.</p> + </item> + <item> + <p>Indexes into counter arrays are one-based. A counter array of + size N contains N counters with index from 1 to N.</p> + </item> + </list> + </description> + + <datatypes> + <datatype> + <name name="counters_ref"/> + <desc><p>Identifies a counter array returned from + <seealso marker="#new/2"><c>new/2</c></seealso>.</p> + </desc> + </datatype> + </datatypes> + + <funcs> + <func> + <name name="new" arity="2"/> + <fsummary>Create counter array</fsummary> + <desc> + <p>Create a new counter array of <c><anno>Size</anno></c> counters.</p> + <p>Argument <c><anno>Opts</anno></c> is a list of the following possible + options:</p> + <taglist> + <tag><c>atomics</c> (Default)</tag> + <item><p>Counters will be sequentially consistent. If write + operation A is done sequencially before write operation B, then a concurrent reader + may see none of them, only A, or both A and B. It cannot see only B.</p> + </item> + <tag><c>write_concurrency</c></tag> + <item><p>This is an optimization to achieve very efficient concurrent + <seealso marker="#add/3"><c>add</c></seealso> and <seealso + marker="#sub/3"><c>sub</c></seealso> operations at the expense of potential read + inconsistency and memory consumption per counter.</p> + <p>Read operations may see sequentially inconsistent results with + regard to concurrent write operations. Even if write operation A is done + sequencially before write operation B, a concurrent reader may see any + combination of A and B, including only B. A read operation is only + guaranteed to see all writes done sequentially before the read. No writes + are ever lost, but will eventually all be seen.</p> + <p>The typical use case for <c>write_concurrency</c> is when + concurrent calls to <seealso marker="#add/3"><c>add</c></seealso> and + <seealso marker="#sub/3"><c>sub</c></seealso> toward the same counters + are very frequent, while calls to <seealso marker="#get/2"><c>get</c> + </seealso> and <seealso marker="#put/3"><c>put</c></seealso> are much + less frequent. The lack of absolute read consistency must also be + acceptable.</p> + </item> + </taglist> + </desc> + </func> + + <func> + <name name="get" arity="2"/> + <fsummary>Read counter value</fsummary> + <desc> + <p>Read counter value.</p> + </desc> + </func> + + <func> + <name name="add" arity="3"/> + <fsummary>Add to counter</fsummary> + <desc> + <p>Add <c><anno>Incr</anno></c> to counter at index + <c><anno>Ix</anno></c>.</p> + </desc> + </func> + + <func> + <name name="sub" arity="3"/> + <fsummary>Subtract from counter</fsummary> + <desc> + <p>Subtract <c><anno>Decr</anno></c> from counter at index + <c><anno>Ix</anno></c>.</p> + </desc> + </func> + + <func> + <name name="put" arity="3"/> + <fsummary>Set counter to value</fsummary> + <desc> + <p>Write <c><anno>Value</anno></c> to counter at index + <c><anno>Ix</anno></c>.</p> + <note> + <p>Despite its name, the <c>write_concurrency</c> optimization does not + improve <c>put</c>. A call to <c>put</c> is a relative heavy + operation compared to the very lightweight and scalable <seealso + marker="#add/3"><c>add</c></seealso> and <seealso marker="#sub/3"> + <c>sub</c></seealso>. The cost for a <c>put</c> with + <c>write_concurrency</c> is lika a <seealso marker="#get/2"><c>get</c> + </seealso> plus a <c>put</c> without <c>write_concurrency</c>.</p> + </note> + </desc> + </func> + + <func> + <name name="info" arity="1"/> + <fsummary>Get information about counter array.</fsummary> + <desc> + <p>Return information about a counter array in a map. The map + has the following keys (at least):</p> + <taglist> + <tag><c>size</c></tag> + <item><p>The number of counters in the array.</p></item> + <tag><c>memory</c></tag> + <item><p>Approximate memory consumption for the array in + bytes.</p></item> + </taglist> + </desc> + </func> + + </funcs> +</erlref> |