Cache backends

A variety of storage backends exists. They have different characteristics and can be used for different caching needs. The best backend depends on a given server setup and hardware, as well as cache type and usage. A backend should be chosen wisely, as a wrong decision could end up actually slowing down a TYPO3 installation.

Backend API

All backends must implement at least interface TYPO3\CMS\Core\Cache\Backend\BackendInterface.

Changed in version 14.0

The \FreezableBackendInterface has been removed. See Breaking: #107310 - Remove FreezableBackendInterface.

BackendInterface

interface BackendInterface

Fully qualified name: \TYPO3\CMS\Core\Cache\Backend\BackendInterface

A contract for a Cache Backend

setCache ( \TYPO3\CMS\Core\Cache\Frontend\FrontendInterface $cache)

Sets a reference to the cache frontend which uses this backend

param $cache: The frontend for this backend

set ( ?string $entryIdentifier, ?string $data, array $tags = [], ?int $lifetime = NULL)

Saves data in the cache.

param $entryIdentifier: An identifier for this specific cache entry
param $data: The data to be stored
param $tags: Tags to associate with this cache entry. If the backend does not support tags, this option can be ignored., default: []
param $lifetime: Lifetime of this cache entry in seconds. If NULL is specified, the default lifetime is used. "0" means unlimited lifetime., default: NULL

get ( ?string $entryIdentifier)

Loads data from the cache.

param $entryIdentifier: An identifier which describes the cache entry to load
Return description: The cache entry's content as a string or FALSE if the cache entry could not be loaded

Returns: mixed

has ( ?string $entryIdentifier)

Checks if a cache entry with the specified identifier exists.

param $entryIdentifier: An identifier specifying the cache entry
Return description: TRUE if such an entry exists, FALSE if not

Returns: bool

remove ( ?string $entryIdentifier)

Removes all cache entries matching the specified identifier.

Usually this only affects one entry but if - for what reason ever - old entries for the identifier still exist, they are removed as well.

param $entryIdentifier: Specifies the cache entry to remove
Return description: TRUE if (at least) an entry could be removed or FALSE if no entry was found

Returns: bool

flush ( ): Removes all cache entries of this cache.

collectGarbage ( ): Does garbage collection

All operations on a specific cache must be done with these methods. There are several further interfaces that can be implemented by backends to declare additional capabilities. Usually, extension code should not handle cache backend operations directly, but should use the frontend object instead.

TaggableBackendInterface

interface TaggableBackendInterface

Fully qualified name: \TYPO3\CMS\Core\Cache\Backend\TaggableBackendInterface

A contract for a cache backend which supports tagging.

flushByTag ( ?string $tag)

Removes all cache entries of this cache which are tagged by the specified tag.

param $tag: The tag the entries must have

flushByTags ( array $tags)

Removes all cache entries of this cache which are tagged by any of the specified tags.

param $tags: List of tags

findIdentifiersByTag ( ?string $tag)

Finds and returns all cache entry identifiers which are tagged by the specified tag

param $tag: The tag to search for
Return description: An array with identifiers of all matching entries. An empty array if no entries matched

Returns: array

setCache ( \TYPO3\CMS\Core\Cache\Frontend\FrontendInterface $cache)

Sets a reference to the cache frontend which uses this backend

param $cache: The frontend for this backend

set ( ?string $entryIdentifier, ?string $data, array $tags = [], ?int $lifetime = NULL)

Saves data in the cache.

param $entryIdentifier: An identifier for this specific cache entry
param $data: The data to be stored
param $tags: Tags to associate with this cache entry. If the backend does not support tags, this option can be ignored., default: []
param $lifetime: Lifetime of this cache entry in seconds. If NULL is specified, the default lifetime is used. "0" means unlimited lifetime., default: NULL

get ( ?string $entryIdentifier)

Loads data from the cache.

param $entryIdentifier: An identifier which describes the cache entry to load
Return description: The cache entry's content as a string or FALSE if the cache entry could not be loaded

Returns: mixed

has ( ?string $entryIdentifier)

Checks if a cache entry with the specified identifier exists.

param $entryIdentifier: An identifier specifying the cache entry
Return description: TRUE if such an entry exists, FALSE if not

Returns: bool

remove ( ?string $entryIdentifier)

Removes all cache entries matching the specified identifier.

Usually this only affects one entry but if - for what reason ever - old entries for the identifier still exist, they are removed as well.

param $entryIdentifier: Specifies the cache entry to remove
Return description: TRUE if (at least) an entry could be removed or FALSE if no entry was found

Returns: bool

flush ( ): Removes all cache entries of this cache.

collectGarbage ( ): Does garbage collection

PhpCapableBackendInterface

interface PhpCapableBackendInterface

Fully qualified name: \TYPO3\CMS\Core\Cache\Backend\PhpCapableBackendInterface

A contract for a cache backend which is capable of storing, retrieving and including PHP source code.

requireOnce ( ?string $entryIdentifier)

Loads PHP code from the cache and require_onces it right away.

param $entryIdentifier: An identifier which describes the cache entry to load
Return description: Potential return value from the include operation

Returns: mixed

setCache ( \TYPO3\CMS\Core\Cache\Frontend\FrontendInterface $cache)

Sets a reference to the cache frontend which uses this backend

param $cache: The frontend for this backend

set ( ?string $entryIdentifier, ?string $data, array $tags = [], ?int $lifetime = NULL)

Saves data in the cache.

param $entryIdentifier: An identifier for this specific cache entry
param $data: The data to be stored
param $tags: Tags to associate with this cache entry. If the backend does not support tags, this option can be ignored., default: []
param $lifetime: Lifetime of this cache entry in seconds. If NULL is specified, the default lifetime is used. "0" means unlimited lifetime., default: NULL

get ( ?string $entryIdentifier)

Loads data from the cache.

param $entryIdentifier: An identifier which describes the cache entry to load
Return description: The cache entry's content as a string or FALSE if the cache entry could not be loaded

Returns: mixed

has ( ?string $entryIdentifier)

Checks if a cache entry with the specified identifier exists.

param $entryIdentifier: An identifier specifying the cache entry
Return description: TRUE if such an entry exists, FALSE if not

Returns: bool

remove ( ?string $entryIdentifier)

Removes all cache entries matching the specified identifier.

Usually this only affects one entry but if - for what reason ever - old entries for the identifier still exist, they are removed as well.

param $entryIdentifier: Specifies the cache entry to remove
Return description: TRUE if (at least) an entry could be removed or FALSE if no entry was found

Returns: bool

flush ( ): Removes all cache entries of this cache.

collectGarbage ( ): Does garbage collection

Common options of caching backends

defaultLifetime

Type: integer
Default: 3600

Default lifetime in seconds of a cache entry if it is not specified for a specific entry on set().

Database Backend

This is the main backend suitable for most storage needs. It does not require additional server daemons nor server configuration.

The database backend does not automatically perform garbage collection. Instead the Scheduler garbage collection task should be used.

It stores data in the configured database (usually MySQL) and can handle large amounts of data with reasonable performance. Data and tags are stored in two different tables, every cache needs its own set of tables. In terms of performance the database backend is already pretty well optimized and should be used as default backend if in doubt. This backend is the default backend if no backend is specifically set in the configuration.

The Core takes care of creating and updating required database tables "on the fly".

Note

However, caching framework tables which are not needed anymore are not deleted automatically. That is why the database analyzer in the Install Tool will propose you to rename/delete caching framework tables after you changed the caching backend to a non-database one.

For caches with a lot of read and write operations, it is important to tune the MySQL setup. The most important setting is innodb_buffer_pool_size. A generic goal is to give MySQL as much RAM as needed to have the main table space loaded completely in memory.

The database backend tends to slow down if there are many write operations and big caches which do not fit into memory because of slow harddrive seek and write performance. If the data table grows too big to fit into memory, it is possible to compress given data transparently with this backend, which often shrinks the amount of needed space to 1/4 or less. The overhead of the compress/uncompress operation is usually not high. A good candidate for a cache with enabled compression is the Core pages cache: it is only read or written once per request and the data size is pretty large. The compression should not be enabled for caches which are read or written multiple times during one request.

InnoDB Issues

The database backend for MySQL uses InnoDB tables. Due to the nature of InnoDB, deleting records does not reclaim the actual disk space. E.g. if the cache uses 10GB, cleaning it will still keep 10GB allocated on the disk even though phpMyAdmin will show 0 as the cache table size. To reclaim the space, turn on the MySQL option file_per_table, drop the cache tables and re-create them using the Install Tool. This does not by any mean that you should skip the scheduler task. Deleting records still improves performance.

Options of database backends

compression

Type: boolean
Default: false

Whether or not data should be compressed with gzip. This can reduce size of the cache data table, but incurs CPU overhead for compression and decompression.

compressionLevel

Type: integer from -1 to 9
Default: -1

Gzip compression level (if the compression option is set to true). The default compression level is usually sufficient.

-1: Default gzip compression (recommended)
0: No compression
9: Maximum compression (costs a lot of CPU)

Memcached backend

Memcached is a simple, distributed key/value RAM database. To use this backend, at least one memcached daemon must be reachable, and the PECL module "memcache" must be loaded. There are two PHP memcached implementations: "memcache" and "memcached". Currently, only memcache is supported by this backend.

Limitations of memcached backends

Memcached is a simple key-value store by design . Since the caching framework needs to structure it to store the identifier-data-tags relations, for each cache entry it stores an identifier->data, identifier->tags and a tag->identifiers entry.

This leads to structural problems:

If memcache runs out of memory but must store new entries, it will toss some other entry out of the cache (this is called an eviction in memcached speak).
If data is shared over multiple memcache servers and some server fails, key/value pairs on this system will just vanish from cache.

Both cases lead to corrupted caches. If, for example, a tags->identifier entry is lost, dropByTag() will not be able to find the corresponding identifier->data entries which should be removed and they will not be deleted. This results in old data delivered by the cache. Additionally, there is currently no implementation of the garbage collection that could rebuild cache integrity.

It is important to monitor a memcached system for evictions and server outages and to clear caches if that happens.

Furthermore memcache has no sort of namespacing. To distinguish entries of multiple caches from each other, every entry is prefixed with the cache name. This can lead to very long runtimes if a big cache needs to be flushed, because every entry has to be handled separately and it is not possible to just truncate the whole cache with one call as this would clear the whole memcached data which might even hold non TYPO3 related entries.

Because of the mentioned drawbacks, the memcached backend should be used with care or in situations where cache integrity is not important or if a cache has no need to use tags at all. Currently, the memcache backend implements the TaggableBackendInterface, so the implementation does allow tagging, even if it is not advised to used this backend together with heavy tagging.

Warning

Since memcached has no sort of namespacing and access control, this backend should not be used if other third party systems have access to the same memcached daemon for security reasons. This is a typical problem in cloud deployments where access to memcache is cheap (but could be read by third parties) and access to databases is expensive.

Options for the memcached backend

servers

Type: array
Required: true

Array of used memcached servers. At least one server must be defined. Each server definition is a string, allowed syntaxes:

hostname or IP: TCP connect to host on memcached default port (usually 11211, defined by PHP ini variable memcache.default_port)
hostname:port: TCP connect to host on port
tcp://hostname:port: Same as above
unix:///path/to/memcached.sock: Connect to memcached server using unix sockets

compression

Type: boolean
Default: false

Enable memcached internal data compression. Can be used to reduce memcached memory consumption, but adds additional compression / decompression CPU overhead on the related memcached servers.

Redis Backend

Redis is a key-value storage/database. In contrast to memcached, it allows structured values. Data is stored in RAM but it allows persistence to disk and doesn't suffer from the design problems of the memcached backend implementation. The redis backend can be used as an alternative to the database backend for big cache tables and helps to reduce load on database servers this way. The implementation can handle millions of cache entries each with hundreds of tags if the underlying server has enough memory.

Redis is known to be extremely fast but very memory hungry. The implementation is an option for big caches with lots of data because most important operations perform O(1) in proportion to the number of (redis) keys. This basically means that the access to an entry in a cache with a million entries is not slower than to a cache with only 10 entries, at least if there is enough memory available to hold the complete set in memory. At the moment only one redis server can be used at a time per cache, but one redis instance can handle multiple caches without performance loss when flushing a single cache.

Attention

The scheduler garbage collection task should be run regularly to find and delete old cache tags entries. These do not expire on their own and would remain in memory indefinitely - unless cache is flushed.

The implementation is based on the PHP phpredis module, which must be available on the system.

Warning

Please check the section on configuration and monitor memory usage (and eviction, if enabled). Otherwise, you may run into problems, if not enough memory for the cache entries is reserved in the Redis server (maxmemory).

Note

It is important to monitor the redis server and tune its settings to the specific caching needs and hardware capabilities. There are several articles on the net and the redis configuration file contains some important hints on how to speed up the system if it reaches bounds. A full documentation of available options is far beyond this documentation.

Redis example

The Redis caching backend configuration is very similar to that of other backends, but there is one caveat.

TYPO3 caches should be separated in case the same keys are used. This applies to the pages and pagesection caches. Both use "tagIdents:pageId_21566" for a page with an id of 21566. How you separate them is more of a system administrator decision. We provide examples with several databases but this may not be the best option in production where you might want to use multiple cores (which do not support databases). The separation has the additional advantage that caches can be flushed individually.

If you have several of your own caches which each use unique keys (for example by using a different prefix for the cache identifier for each cache), you can store them in the same database, but it is good practice to separate the core caches.

In practical terms, Redis databases should be used to separate different keys belonging to the same application (if needed), and not to use a single Redis instance for multiple unrelated applications.

https://redis.io/commands/select/

config/system/additional.php | typo3conf/system/additional.php

<?php

$redisHost = '127.0.0.1';
$redisPort = 6379;
$redisCaches = [
    'pages' => [
        'defaultLifetime' => 86400 * 7, // 1 week
        'compression' => true,
    ],
    'pagesection' => [
        'defaultLifetime' => 86400 * 7,
    ],
    'hash' => [],
    'rootline' => [],
];

$redisDatabase = 0;
foreach ($redisCaches as $name => $values) {
    $GLOBALS['TYPO3_CONF_VARS']['SYS']['caching']['cacheConfigurations'][$name]['backend']
        = \TYPO3\CMS\Core\Cache\Backend\RedisBackend::class;
    $GLOBALS['TYPO3_CONF_VARS']['SYS']['caching']['cacheConfigurations'][$name]['options'] = [
        'database' => $redisDatabase++,
        'hostname' => $redisHost,
        'port' => $redisPort,
    ];
    if (isset($values['defaultLifetime'])) {
        $GLOBALS['TYPO3_CONF_VARS']['SYS']['caching']['cacheConfigurations'][$name]['options']['defaultLifetime']
            = $values['defaultLifetime'];
    }
    if (isset($values['compression'])) {
        $GLOBALS['TYPO3_CONF_VARS']['SYS']['caching']['cacheConfigurations'][$name]['options']['compression']
            = $values['compression'];
    }
}

Options for the redis caching backend

servers

Type: string
Default: 127.0.0.1

IP address or name of redis server to connect to.

port

Type: integer
Default: 6379

Port of the redis daemon.

persistentConnection

Type: boolean
Default: false

Activate a persistent connection to redis server. This could be a benefit under high load cloud setups.

database

Type: integer
Default: 0

Number of the database to store entries. Each cache should use its own database, otherwise all caches sharing a database are flushed if the flush operation is issued to one of them. Database numbers 0 and 1 are used and flushed by the Core unit tests and should not be used if possible.

password

Type: string

Password used to connect to the redis instance if the redis server needs authentication.

Warning

The password is sent to the redis server in plain text.

compression

Type: boolean
Default: false

Whether or not data compression with gzip should be enabled. This can reduce cache size, but adds some CPU overhead for the compression and decompression operations in PHP.

compressionLevel

Type: integer from -1 to 9
Default: -1

Set gzip compression level to a specific value. The default compression level is usually sufficient.

-1: Default gzip compression (recommended)
0: No compression
9: Maximum compression (but more CPU overhead)

Redis server configuration

This section is about the configuration on the Redis server, not the client.

For the flushing by cache tags to work, it is important that the integrity of the cache entries and cache tags is maintained. This may not be the case, depending on which eviction policy (maxmemory-policy) is used. For example, for a page id=81712, the following entries may exist in the Redis page cache:

tagIdents:pageId_81712 (tag->identifier relation)
identTags:81712_7e9c8309692aa221b08e6d5f6ec09fb6 (identifier->tags relation)
identData:81712_7e9c8309692aa221b08e6d5f6ec09fb6 (identifier->data)

If entries are evicted (due to memory shortage), there is no mechanism in place which ensures that all entries which are related, will be evicted. If maxmemory-policy allkeys-lru is used, for example, this may result in the situation that the cache entry (identData) still exists, but the tag entry (tagIdents) does not. The tag entry reflects the relation "cache tag => cache identifier" and is used for RedisBackend::flushByTag()). If this entry is gone, the cache can no longer be flushed if content is changed on the page or an explicit flushing of the page cache for this page is requested. Once this is the case, cache flushing (for this page) is only possible via other means (such as full cache flush).

Because of this, the following recommendations apply:

Allocate enough memory (maxmemory) for the cache.
Use the maxmemory-policy volatile-ttl. This will ensure that no tagIdents entries are removed. (These have no expiration date).
Regularly run the TYPO3 scheduler garbage collection task for the Redis cache backend.
Monitor evicted_keys in case an eviction policy is used.
Monitor used_memory if eviction policy noeviction is used. The used_memory should always be less then maxmemory.

Tip

The information about evicted_keys etc. can be obtained via redis-cli and the info command or via php-redis. Further information of the results of info is in the documentation.

The maxmemory-policy options have the following drawbacks:

volatile-ttl: (recommended) Will flush only entries with an expiration date. Should be ok with TYPO3.
noeviction: (Not recommended) Once memory is full, no new entries will be saved to cache. Only use if you can ensure that there is always enough memory.
allkeys-lru, allkeys-lfu, allkeys-random: (Not recommended) This may result in tagIdents being removed, but not the related identData entry, which makes it impossible to flush the cache entries by tag (which is necessary for TYPO3 cache flushing on changes to work and the flush page cache to work for specific pages).

File backend

The file backend stores every cache entry as a single file to the file system. The lifetime and tags are added after the data part in the same file.

This backend is the big brother of the Simple file backend and implements additional interfaces. Like the simple file backend it also implements the PhpCapableInterface, so it can be used with PhpFrontend. In contrast to the simple file backend it furthermore implements TaggableInterface.

In general, the backend was specifically optimized to cache PHP code, the get and set operations have low overhead. The file backend is not very good with tagging and does not scale well with the number of tags. Do not use this backend if cached data has many tags.

Warning

The performance of flushByTag() is bad and scales just O(n).

On the contrary performance of get() and set() operations. is good and scales well. Of course if many entries have to be handled, this might still slow down after a while and a different storage strategy should be used (e.g. RAM disks, battery backed up RAID systems or SSD hard disks).

Options for the file backend

cacheDirectory

Type: array
Default: var/cache/

The directory where the cache files are stored. By default it is assumed that the directory is below TYPO3_DOCUMENT_ROOT. However, an absolute path can be selected, too. Every cache should be assigned its own directory, otherwise flushing of one cache would flush all other caches within the same directory as well.

Simple File Backend

The simple file backend is the small brother of the file backend. In contrast to most other backends, it does not implement the TaggableInterface, so cache entries can not be tagged and flushed by tag. This improves the performance if cache entries do not need such tagging. The TYPO3 Core uses this backend for its central Core cache (that hold autoloader cache entries and other important cache entries). The Core cache is usually flushed completely and does not need specific cache entry eviction.

PDO Backend

The PDO backend can be used as a native PDO interface to databases which are connected to PHP via PDO. It is an alternative to the database backend if a cache should be stored in a database which is otherwise only supported by TYPO3 dbal to reduce the parser overhead.

The garbage collection is implemented for this backend and should be called to clean up hard disk space or memory.

Note

There is currently very little production experience with this backend, especially not with a capable database like Oracle. Any feedback for real life use cases of this cache is appreciated.

Options for the PDO backend

dataSourceName

Type: string
Required: true

Data source name for connecting to the database. Examples:

mysql:host=localhost;dbname=test
sqlite:/path/to/sqlite.db
sqlite::memory

username

Type: string

Username for the database connection.

password

Type: string

Password to use for the database connection.

Transient Memory Backend

The transient memory backend stores data in a PHP array. It is only valid for one request. This becomes handy if code logic needs to do expensive calculations or must look up identical information from a database over and over again during its execution. In this case it is useful to store the data in an array once and lookup the entry from the cache for consecutive calls to get rid of the otherwise additional overhead. Since caches are available system wide and shared between Core and extensions they can profit from each other if they need the same information.

Since the data is stored directly in memory, this backend is the quickest backend available. The stored data adds to the memory consumed by the PHP process and can hit the memory_limit PHP setting.

Null Backend

The null backend is a dummy backend which doesn't store any data and always returns false on get(). This backend becomes handy in development context to practically "switch off" a cache.