diff options
author | Björn Gustavsson <[email protected]> | 2018-11-06 11:56:50 +0100 |
---|---|---|
committer | Björn Gustavsson <[email protected]> | 2018-11-06 11:56:50 +0100 |
commit | adffe592c98149ac3b7333cf6e46970a13a57350 (patch) | |
tree | c22ccbe14e456bb0da7547ada016e3b1fef4fdd0 /erts/doc/src/persistent_term.xml | |
parent | e7e3f59e5350bb39f57f0403f3c78443c58039b1 (diff) | |
parent | 7489e818de5da5e1648ac06356eeefdc98902564 (diff) | |
download | otp-adffe592c98149ac3b7333cf6e46970a13a57350.tar.gz otp-adffe592c98149ac3b7333cf6e46970a13a57350.tar.bz2 otp-adffe592c98149ac3b7333cf6e46970a13a57350.zip |
Merge branch 'bjorn/erts/persistent_terms/OTP-14669' into maint
* bjorn/erts/persistent_terms/OTP-14669:
Implement a tab for persistent terms in crashdump viewer
Add tests of persistent terms for crashdump_viewer
Add a persistent term storage
Refactor releasing of literals
Extend the sharing-preserving routines to optionally copy literals
Diffstat (limited to 'erts/doc/src/persistent_term.xml')
-rw-r--r-- | erts/doc/src/persistent_term.xml | 290 |
1 files changed, 290 insertions, 0 deletions
diff --git a/erts/doc/src/persistent_term.xml b/erts/doc/src/persistent_term.xml new file mode 100644 index 0000000000..d2a138d65f --- /dev/null +++ b/erts/doc/src/persistent_term.xml @@ -0,0 +1,290 @@ +<?xml version="1.0" encoding="utf-8" ?> +<!DOCTYPE erlref SYSTEM "erlref.dtd"> + +<erlref> + <header> + <copyright> + <year>2018</year><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>persistent_term</title> + <prepared></prepared> + <docno></docno> + <date></date> + <rev></rev> + <file>persistent_term.xml</file> + </header> + <module>persistent_term</module> + <modulesummary>Persistent terms.</modulesummary> + <description> + <p>This module is similar to <seealso + marker="stdlib:ets"><c>ets</c></seealso> in that it provides a + storage for Erlang terms that can be accessed in constant time, + but with the difference that <c>persistent_term</c> has been + highly optimized for reading terms at the expense of writing and + updating terms. When a persistent term is updated or deleted, a + global garbage collection pass is run to scan all processes for + the deleted term, and to copy it into each process that still uses + it. Therefore, <c>persistent_term</c> is suitable for storing + Erlang terms that are frequently accessed but never or + infrequently updated.</p> + + <warning><p>Persistent terms is an advanced feature and is not a + general replacement for ETS tables. Before using persistent terms, + make sure to fully understand the consequence to system + performance when updating or deleting persistent terms.</p></warning> + + <p>Term lookup (using <seealso + marker="#get/1"><c>get/1</c></seealso>), is done in constant time + and without taking any locks, and the term is <strong>not</strong> + copied to the heap (as is the case with terms stored in ETS + tables).</p> + + <p>Storing or updating a term (using <seealso + marker="#put/2"><c>put/2</c></seealso>) is proportional to the + number of already created persistent terms because the hash table + holding the keys will be copied. In addition, the term itself will + be copied.</p> + + <p>When a (complex) term is deleted (using <seealso + marker="#erase/1"><c>erase/1</c></seealso>) or replaced by another + (using <seealso marker="#put/2"><c>put/2</c></seealso>), a global + garbage collection is initiated. It works like this:</p> + + <list> + <item><p>All processes in the system will be scheduled to run a + scan of their heaps for the term that has been deleted. While + such scan is relatively light-weight, if there are many + processes, the system can become less responsive until all + process have scanned theirs heaps.</p></item> + + <item><p>If the deleted term (or any part of it) is still used + by a process, that process will do a major (fullsweep) garbage + collection and copy the term into the process. However, at most + two processes at a time will be scheduled to do that kind of + garbage collection.</p></item> + </list> + + <p>Deletion of atoms and other terms that fit in one machine word + is specially optimized to avoid doing a global GC. It is still not + recommended to update persistent terms with such values too + frequently because the hash table holding the keys is copied every + time a persistent term is updated.</p> + + <p>Some examples are suitable uses for persistent terms are:</p> + + <list> + <item><p>Storing of configuration data that must be easily + accessible by all processes.</p></item> + + <item><p>Storing of references for NIF resources.</p></item> + + <item><p>Storing of references for efficient counters.</p></item> + + <item><p>Storing an atom to indicate a logging level or whether debugging + is turned on.</p></item> + </list> + + </description> + + <section> + <title>Storing Huge Persistent Terms</title> + <p>The current implementation of persistent terms uses the literal + <seealso marker="erts_alloc">allocator</seealso> also used for + literals (constant terms) in BEAM code. By default, 1 GB of + virtual address space is reserved for literals in BEAM code and + persistent terms. The amount of virtual address space reserved for + literals can be changed by using the <seealso + marker="erts_alloc#MIscs"><c>+MIscs option</c></seealso> when + starting the emulator.</p> + + <p>Here is an example how the reserved virtual address space for literals + can be raised to 2 GB (2048 MB):</p> + + <pre> + erl +MIscs 2048</pre> + </section> + + <section> + <title>Warning For Many Persistent Terms</title> + <p>The runtime system will send a warning report to the + error logger if more than 20000 persistent terms have been + created. It will look like this:</p> + +<pre> +More than 20000 persistent terms have been created. +It is recommended to avoid creating an excessive number of +persistent terms, as creation and deletion of persistent terms +will be slower as the number of persistent terms increases.</pre> + </section> + + <section> + <title>Best Practices for Using Persistent Terms</title> + + <p>It is recommended to use keys like <c>?MODULE</c> or + <c>{?MODULE,SubKey}</c> to avoid name collisions.</p> + + <p>Prefer creating a few large persistent terms to creating many + small persistent terms. The execution time for storing a + persistent term is proportional to the number of already existing + terms.</p> + + <p>Updating a persistent term with the same value as it already + has is specially optimized to do nothing quickly; thus, there is + no need compare the old and new values and avoid calling + <seealso marker="#put/2"><c>put/2</c></seealso> if the values + are equal.</p> + + <p>When atoms or other terms that fit in one machine word are + deleted, no global GC is needed. Therefore, persistent terms that + have atoms as their values can be updated more frequently, but + note that updating such persistent terms is still much more + expensive than reading them.</p> + + <p>Updating or deleting a persistent term will trigger a global GC + if the term does not fit in one machine word. Processes will be + scheduled as usual, but all processes will be made runnable at + once, which will make the system less responsive until all process + have run and scanned their heaps for the deleted terms. One way to + minimize the effects on responsiveness could be to minimize the + number of processes on the node before updating or deleting a + persistent term. It would also be wise to avoid updating terms + when the system is at peak load.</p> + + <p>Avoid storing a retrieved persistent term in a process if that + persistent term could be deleted or updated in the future. If a + process holds a reference to a persistent term when the term is + deleted, the process will be garbage collected and the term copied + to process.</p> + + <p>Avoid updating or deleting more than one persistent term at a + time. Each deleted term will trigger its own global GC. That + means that deleting N terms will make the system less responsive N + times longer than deleting a single persistent term. Therefore, + terms that are to be updated at the same time should be collected + into a larger term, for example, a map or a tuple.</p> + </section> + + <section> + <title>Example</title> + + <p>The following example shows how lock contention for ETS tables + can be minimized by having one ETS table for each scheduler. The + table identifiers for the ETS tables are stored as a single + persistent term:</p> + +<pre> + %% There is one ETS table for each scheduler. + Sid = erlang:system_info(scheduler_id), + Tid = element(Sid, persistent_term:get(?MODULE)), + ets:update_counter(Tid, Key, 1).</pre> + + </section> + + <datatypes> + <datatype> + <name name="key"/> + <desc> + <p>Any Erlang term.</p> + </desc> + </datatype> + <datatype> + <name name="value"/> + <desc> + <p>Any Erlang term.</p> + </desc> + </datatype> + </datatypes> + + <funcs> + <func> + <name name="erase" arity="1"/> + <fsummary>Erase the name for a persistent term.</fsummary> + <desc> + <p>Erase the name for the persistent term with key + <c><anno>Key</anno></c>. The return value will be <c>true</c> + if there was a persistent term with the key + <c><anno>Key</anno></c>, and <c>false</c> if there was no + persistent term associated with the key.</p> + <p>If there existed a previous persistent term associated with + key <c><anno>Key</anno></c>, a global GC has been initiated + when <c>erase/1</c> returns. See <seealso + marker="#description">Description</seealso>.</p> + </desc> + </func> + + <func> + <name name="get" arity="0"/> + <fsummary>Get all persistent terms.</fsummary> + <desc> + <p>Retrieve the keys and values for all persistent terms. + The keys will be copied to the heap for the process calling + <c>get/0</c>, but the values will not.</p> + </desc> + </func> + + <func> + <name name="get" arity="1"/> + <fsummary>Get the value for a persistent term.</fsummary> + <desc> + <p>Retrieve the value for the persistent term associated with + the key <c><anno>Key</anno></c>. The lookup will be made in + constant time and the value will not be copied to the heap + of the calling process.</p> + <p>This function fails with a <c>badarg</c> exception if no + term has been stored with the key + <c><anno>Key</anno></c>.</p> + <p>If the calling process holds on to the value of the + persistent term and the persistent term is deleted in the future, + the term will be copied to the process.</p> + </desc> + </func> + + <func> + <name name="info" arity="0"/> + <fsummary>Get information about persistent terms.</fsummary> + <desc> + <p>Return information about persistent terms in a map. The map + has the following keys:</p> + <taglist> + <tag><c>count</c></tag> + <item><p>The number of persistent terms.</p></item> + <tag><c>memory</c></tag> + <item><p>The total amount of memory (measured in bytes) + used by all persistent terms.</p></item> + </taglist> + </desc> + </func> + + <func> + <name name="put" arity="2"/> + <fsummary>Store a term.</fsummary> + <desc> + <p>Store the value <c><anno>Value</anno></c> as a persistent term and + associate it with the key <c><anno>Key</anno></c>.</p> + <p>If the value <c><anno>Value</anno></c> is equal to the value + previously stored for the key, <c>put/2</c> will do nothing and return + quickly.</p> + <p>If there existed a previous persistent term associated with + key <c><anno>Key</anno></c>, a global GC has been initiated + when <c>put/2</c> returns. See <seealso + marker="#description">Description</seealso>.</p> + </desc> + </func> + </funcs> +</erlref> |