原文地址 www.redis.net.cn

Redis 配置

Redis 的配置文件位于 Redis 安装目录下，文件名为 redis.conf。

你可以通过 CONFIG 命令查看或设置配置项。

语法

Redis CONFIG 命令格式如下：

  1. redis 127.0.0.1:6379> CONFIG GET CONFIG_SETTING_NAME

实例

  1. redis 127.0.0.1:6379> CONFIG GET loglevel
  2. 1) "loglevel"
  3. 2) "notice"

使用 * 号获取所有配置项：

实例

  1. redis 127.0.0.1:6379> CONFIG GET *
  2.   1) "dbfilename"
  3.   2) "dump.rdb"
  4.   3) "requirepass"
  5.   4) ""
  6.   5) "masterauth"
  7.   6) ""
  8.   7) "unixsocket"
  9.   8) ""
  10.   9) "logfile"
  11.  10) ""
  12.  11) "pidfile"
  13.  12) "/var/run/redis.pid"
  14.  13) "maxmemory"
  15.  14) "0"
  16.  15) "maxmemory-samples"
  17.  16) "3"
  18.  17) "timeout"
  19.  18) "0"
  20.  19) "tcp-keepalive"
  21.  20) "0"
  22.  21) "auto-aof-rewrite-percentage"
  23.  22) "100"
  24.  23) "auto-aof-rewrite-min-size"
  25.  24) "67108864"
  26.  25) "hash-max-ziplist-entries"
  27.  26) "512"
  28.  27) "hash-max-ziplist-value"
  29.  28) "64"
  30.  29) "list-max-ziplist-entries"
  31.  30) "512"
  32.  31) "list-max-ziplist-value"
  33.  32) "64"
  34.  33) "set-max-intset-entries"
  35.  34) "512"
  36.  35) "zset-max-ziplist-entries"
  37.  36) "128"
  38.  37) "zset-max-ziplist-value"
  39.  38) "64"
  40.  39) "hll-sparse-max-bytes"
  41.  40) "3000"
  42.  41) "lua-time-limit"
  43.  42) "5000"
  44.  43) "slowlog-log-slower-than"
  45.  44) "10000"
  46.  45) "latency-monitor-threshold"
  47.  46) "0"
  48.  47) "slowlog-max-len"
  49.  48) "128"
  50.  49) "port"
  51.  50) "6379"
  52.  51) "tcp-backlog"
  53.  52) "511"
  54.  53) "databases"
  55.  54) "16"
  56.  55) "repl-ping-slave-period"
  57.  56) "10"
  58.  57) "repl-timeout"
  59.  58) "60"
  60.  59) "repl-backlog-size"
  61.  60) "1048576"
  62.  61) "repl-backlog-ttl"
  63.  62) "3600"
  64.  63) "maxclients"
  65.  64) "4064"
  66.  65) "watchdog-period"
  67.  66) "0"
  68.  67) "slave-priority"
  69.  68) "100"
  70.  69) "min-slaves-to-write"
  71.  70) "0"
  72.  71) "min-slaves-max-lag"
  73.  72) "10"
  74.  73) "hz"
  75.  74) "10"
  76.  75) "no-appendfsync-on-rewrite"
  77.  76) "no"
  78.  77) "slave-serve-stale-data"
  79.  78) "yes"
  80.  79) "slave-read-only"
  81.  80) "yes"
  82.  81) "stop-writes-on-bgsave-error"
  83.  82) "yes"
  84.  83) "daemonize"
  85.  84) "no"
  86.  85) "rdbcompression"
  87.  86) "yes"
  88.  87) "rdbchecksum"
  89.  88) "yes"
  90.  89) "activerehashing"
  91.  90) "yes"
  92.  91) "repl-disable-tcp-nodelay"
  93.  92) "no"
  94.  93) "aof-rewrite-incremental-fsync"
  95.  94) "yes"
  96.  95) "appendonly"
  97.  96) "no"
  98.  97) "dir"
  99.  98) "/home/deepak/Downloads/redis-2.8.13/src"
  100.  99) "maxmemory-policy"
  101. 100) "volatile-lru"
  102. 101) "appendfsync"
  103. 102) "everysec"
  104. 103) "save"
  105. 104) "3600 1 300 100 60 10000"
  106. 105) "loglevel"
  107. 106) "notice"
  108. 107) "client-output-buffer-limit"
  109. 108) "normal 0 0 0 slave 268435456 67108864 60 pubsub 33554432 8388608 60"
  110. 109) "unixsocketperm"
  111. 110) "0"
  112. 111) "slaveof"
  113. 112) ""
  114. 113) "notify-keyspace-events"
  115. 114) ""
  116. 115) "bind"
  117. 116) ""

编辑配置

你可以通过修改 redis.conf 文件或使用 CONFIG set 命令来修改配置。

语法

CONFIG SET 命令基本语法：

  1. redis 127.0.0.1:6379> CONFIG SET CONFIG_SETTING_NAME NEW_CONFIG_VALUE

实例

  1.  redis 127.0.0.1:6379> CONFIG SET loglevel "notice"
  2.  OK
  3. redis 127.0.0.1:6379> CONFIG GET loglevel
  4. 1) "loglevel"
  5. 2) "notice"

参数说明

redis.conf 配置项说明如下：

  1. Redis默认不是以守护进程的方式运行，可以通过该配置项修改，使用yes启用守护进程

daemonize no

  1. 当Redis以守护进程方式运行时，Redis默认会把pid写入/var/run/redis.pid文件，可以通过pidfile指定

pidfile /var/run/redis.pid

  1. 指定Redis监听端口，默认端口为6379，作者在自己的一篇博文中解释了为什么选用6379作为默认端口，因为6379在手机按键上MERZ对应的号码，而MERZ取自意大利歌女Alessia Merz的名字

port 6379

  1. 绑定的主机地址

bind 127.0.0.1

5.当 客户端闲置多长时间后关闭连接，如果指定为0，表示关闭该功能

timeout 300

  1. 指定日志记录级别，Redis总共支持四个级别：debug、verbose、notice、warning，默认为verbose

loglevel verbose

  1. 日志记录方式，默认为标准输出，如果配置Redis为守护进程方式运行，而这里又配置为日志记录方式为标准输出，则日志将会发送给/dev/null

logfile stdout

  1. 设置数据库的数量，默认数据库为0，可以使用SELECT <dbid>命令在连接上指定数据库id

databases 16

  1. 指定在多长时间内，有多少次更新操作，就将数据同步到数据文件，可以多个条件配合

save <seconds> <changes>

Redis默认配置文件中提供了三个条件：

save 900 1

  1. **save 300 10**
  3. **save 60 10000**

分别表示900秒（15分钟）内有1个更改，300秒（5分钟）内有10个更改以及60秒内有10000个更改。

  1. 指定存储至本地数据库时是否压缩数据，默认为yes，Redis采用LZF压缩，如果为了节省CPU时间，可以关闭该选项，但会导致数据库文件变的巨大

rdbcompression yes

  1. 指定本地数据库文件名，默认值为dump.rdb

dbfilename dump.rdb

  1. 指定本地数据库存放目录

dir ./

  1. 设置当本机为slav服务时，设置master服务的IP地址及端口，在Redis启动时，它会自动从master进行数据同步

slaveof <masterip> <masterport>

  1. 当master服务设置了密码保护时，slav服务连接master的密码

masterauth <master-password>

  1. 设置Redis连接密码，如果配置了连接密码，客户端在连接Redis时需要通过AUTH <password>命令提供密码，默认关闭

requirepass foobared

  1. 设置同一时间最大客户端连接数，默认无限制，Redis可以同时打开的客户端连接数为Redis进程可以打开的最大文件描述符数，如果设置 maxclients 0，表示不作限制。当客户端连接数到达限制时，Redis会关闭新的连接并向客户端返回max number of clients reached错误信息

maxclients 128

  1. 指定Redis最大内存限制，Redis在启动时会把数据加载到内存中，达到最大内存后，Redis会先尝试清除已到期或即将到期的Key，当此方法处理 后，仍然到达最大内存设置，将无法再进行写入操作，但仍然可以进行读取操作。Redis新的vm机制，会把Key存放内存，Value会存放在swap区

maxmemory <bytes>

  1. 指定是否在每次更新操作后进行日志记录，Redis在默认情况下是异步的把数据写入磁盘，如果不开启，可能会在断电时导致一段时间内的数据丢失。因为 redis本身同步数据文件是按上面save条件来同步的，所以有的数据会在一段时间内只存在于内存中。默认为no

appendonly no

  1. 指定更新日志文件名，默认为appendonly.aof

appendfilename appendonly.aof

  1. 指定更新日志条件，共有3个可选值： no：表示等操作系统进行数据缓存同步到磁盘（快） always：表示每次更新操作后手动调用fsync()将数据写到磁盘（慢，安全） everysec：表示每秒同步一次（折衷，默认值）

appendfsync everysec

  1. 指定是否启用虚拟内存机制，默认值为no，简单的介绍一下，VM机制将数据分页存放，由Redis将访问量较少的页即冷数据swap到磁盘上，访问多的页面由磁盘自动换出到内存中（在后面的文章我会仔细分析Redis的VM机制）

vm-enabled no

  1. 虚拟内存文件路径，默认值为/tmp/redis.swap，不可多个Redis实例共享

vm-swap-file /tmp/redis.swap

  1. 将所有大于vm-max-memory的数据存入虚拟内存,无论vm-max-memory设置多小,所有索引数据都是内存存储的(Redis的索引数据 就是keys),也就是说,当vm-max-memory设置为0的时候,其实是所有value都存在于磁盘。默认值为0

vm-max-memory 0

  1. Redis swap文件分成了很多的page，一个对象可以保存在多个page上面，但一个page上不能被多个对象共享，vm-page-size是要根据存储的 数据大小来设定的，作者建议如果存储很多小对象，page大小最好设置为32或者64bytes；如果存储很大大对象，则可以使用更大的page，如果不 确定，就使用默认值

vm-page-size 32

  1. 设置swap文件中的page数量，由于页表（一种表示页面空闲或使用的bitmap）是在放在内存中的，，在磁盘上每8个pages将消耗1byte的内存。

vm-pages 134217728

  1. 设置访问swap文件的线程数,最好不要超过机器的核数,如果设置为0,那么所有对swap文件的操作都是串行的，可能会造成比较长时间的延迟。默认值为4

vm-max-threads 4

  1. 设置在向客户端应答时，是否把较小的包合并为一个包发送，默认为开启

glueoutputbuf yes

  1. 指定在超过一定的数量或者最大的元素超过某一临界值时，采用一种特殊的哈希算法

hash-max-zipmap-entries 64

  1. **hash-max-zipmap-value 512**
  1. 指定是否激活重置哈希，默认为开启（后面在介绍Redis的哈希算法时具体介绍）

activerehashing yes

  1. 指定包含其它的配置文件，可以在同一主机上多个Redis实例之间使用同一份配置文件，而同时各个实例又拥有自己的特定配置文件

include /path/to/local.conf

redis配置文件redis.conf介绍

  1. ###Units单位###
    配置大小单位,开头定义了一些基本的度量单位，只支持bytes，不支持bit
    大小写不敏感
    Redis 配置 - 图1

  2. ###INCLUDES包含###
    Redis 配置 - 图2
    类似jsp中的include，多实例的情况可以把公用的配置文件提取出来

  3. ###网络相关配置 ###
    1.bind
    默认情况bind=127.0.0.1只能接受本机的访问请求
    不写的情况下，无限制接受任何ip地址的访问
    生产环境肯定要写你应用服务器的地址；服务器是需要远程访问的，所以需要将其注释掉
    如果开启了protected-mode，那么在没有设定bind ip且没有设密码的情况下，Redis只允许接受本机的响应
    Redis 配置 - 图3
    保存配置，停止服务，重启启动查看进程，不再是本机访问了。
    Redis 配置 - 图4

    1. protected-mode
      将本机访问保护模式设置no
      Redis 配置 - 图5
  1. 端口号，默认 6379

  1. tcp-backlog
    设置tcp的backlog，backlog其实是一个连接队列，backlog队列总和=未完成三次握手队列+ 已经完成三次握手队列。
    在高并发环境下你需要一个高backlog值来避免慢客户端连接问题。
    注意Linux内核会将这个值减小到/proc/sys/net/core/somaxconn的值（128），所以需要确认增大/proc/sys/net/core/somaxconn和/proc/sys/net/ipv4/tcp_max_syn_backlog（128）两个值来达到想要的效果
    Redis 配置 - 图6

  2. timeout
    一个空闲的客户端维持多少秒会关闭，0表示关闭该功能。即永不关闭。
    Redis 配置 - 图7

  3. tcp-keepalive
    对访问客户端的一种心跳检测，每个n秒检测一次。
    单位为秒，如果设置为0，则不会进行Keepalive检测，建议设置成60
    Redis 配置 - 图8

  1. ###GENERAL通用###

    1. daemonize

Redis 配置 - 图9

  1. pidfile
    存放pid文件的位置，每个实例会产生一个不同的pid文件
    Redis 配置 - 图10

  2. loglevel
    指定日志记录级别，Redis总共支持四个级别：debug、verbose、notice、warning，默认为notice
    四个级别根据使用阶段来选择，生产环境选择notice 或者warning
    Redis 配置 - 图11

  3. logfile
    日志文件名称
    Redis 配置 - 图12

  4. databases 16
    设定库的数量 默认16，默认数据库为0，可以使用SELECT
    命令在连接上指定数据库id
    Redis 配置 - 图13

  1. ###SECURITY安全###

    1. 设置密码
      Redis 配置 - 图14
      访问密码的查看、设置和取消
      在命令中设置密码，只是临时的。重启redis服务器，密码就还原了。
      永久设置，需要再配置文件中进行设置。
      Redis 配置 - 图15

  2. #### LIMITS限制 ###

    1. maxclients

      • 设置redis同时可以与多少个客户端进行连接。

      • 默认情况下为10000个客户端。

      • 如果达到了此限制，redis则会拒绝新的连接请求，并且向这些连接请求方发出“max number of clients reached”以作回应。

Redis 配置 - 图16

  1. maxmemory

    • 建议必须设置，否则，将内存占满，造成服务器宕机

    • 设置redis可以使用的内存量。一旦到达内存使用上限，redis将会试图移除内部数据，移除规则可以通过maxmemory-policy来指定。

    • 如果redis无法根据移除规则来移除内存中的数据，或者设置了“不允许移除”，那么redis则会针对那些需要申请内存的指令返回错误信息，比如SET、LPUSH等。

    • 但是对于无内存申请的指令，仍然会正常响应，比如GET等。如果你的redis是主redis（说明你的redis有从redis），那么在设置内存使用上限时，需要在系统中留出一些内存空间给同步队列缓存，只有在你设置的是“不移除”的情况下，才不用考虑这个因素。

Redis 配置 - 图17

  1. maxmemory-policy

    • volatile-lru：使用LRU算法移除key，只对设置了过期时间的键；（最近最少使用）

    • allkeys-lru：在所有集合key中，使用LRU算法移除key

    • volatile-random：在过期集合中移除随机的key，只对设置了过期时间的键

    • allkeys-random：在所有集合key中，移除随机的key

    • volatile-ttl：移除那些TTL值最小的key，即那些最近要过期的key

    • noeviction：不进行移除。针对写操作，只是返回错误信息

Redis 配置 - 图18

  1. maxmemory-samples

    • 设置样本数量，LRU算法和最小TTL算法都并非是精确的算法，而是估算值，所以你可以设置样本的大小，redis默认会检查这么多个key并选择其中LRU的那个。

    • 一般设置3到7的数字，数值越小样本越不准确，但性能消耗越小。

Redis 配置 - 图19

redis 6.2.3 redis-conf 文件

  1. # Redis configuration file example.
  2. #
  3. # Note that in order to read the configuration file, Redis must be
  4. # started with the file path as first argument:
  5. #
  6. # ./redis-server /path/to/redis.conf
  8. # Note on units: when memory size is needed, it is possible to specify
  9. # it in the usual form of 1k 5GB 4M and so forth:
  10. #
  11. # 1k => 1000 bytes
  12. # 1kb => 1024 bytes
  13. # 1m => 1000000 bytes
  14. # 1mb => 1024*1024 bytes
  15. # 1g => 1000000000 bytes
  16. # 1gb => 1024*1024*1024 bytes
  17. #
  18. # units are case insensitive so 1GB 1Gb 1gB are all the same.
  20. ################################## INCLUDES ###################################
  22. # Include one or more other config files here.  This is useful if you
  23. # have a standard template that goes to all Redis servers but also need
  24. # to customize a few per-server settings.  Include files can include
  25. # other files, so use this wisely.
  26. #
  27. # Note that option "include" won't be rewritten by command "CONFIG REWRITE"
  28. # from admin or Redis Sentinel. Since Redis always uses the last processed
  29. # line as value of a configuration directive, you'd better put includes
  30. # at the beginning of this file to avoid overwriting config change at runtime.
  31. #
  32. # If instead you are interested in using includes to override configuration
  33. # options, it is better to use include as the last line.
  34. #
  35. # include /path/to/local.conf
  36. # include /path/to/other.conf
  38. ################################## MODULES #####################################
  40. # Load modules at startup. If the server is not able to load modules
  41. # it will abort. It is possible to use multiple loadmodule directives.
  42. #
  43. # loadmodule /path/to/my_module.so
  44. # loadmodule /path/to/other_module.so
  46. ################################## NETWORK #####################################
  48. # By default, if no "bind" configuration directive is specified, Redis listens
  49. # for connections from all available network interfaces on the host machine.
  50. # It is possible to listen to just one or multiple selected interfaces using
  51. # the "bind" configuration directive, followed by one or more IP addresses.
  52. # Each address can be prefixed by "-", which means that redis will not fail to
  53. # start if the address is not available. Being not available only refers to
  54. # addresses that does not correspond to any network interfece. Addresses that
  55. # are already in use will always fail, and unsupported protocols will always BE
  56. # silently skipped.
  57. #
  58. # Examples:
  59. #
  60. # bind 192.168.1.100 10.0.0.1     # listens on two specific IPv4 addresses
  61. # bind 127.0.0.1 ::1              # listens on loopback IPv4 and IPv6
  62. # bind * -::*                     # like the default, all available interfaces
  63. #
  64. # ~~~ WARNING ~~~ If the computer running Redis is directly exposed to the
  65. # internet, binding to all the interfaces is dangerous and will expose the
  66. # instance to everybody on the internet. So by default we uncomment the
  67. # following bind directive, that will force Redis to listen only on the
  68. # IPv4 and IPv6 (if available) loopback interface addresses (this means Redis
  69. # will only be able to accept client connections from the same host that it is
  70. # running on).
  71. #
  72. # IF YOU ARE SURE YOU WANT YOUR INSTANCE TO LISTEN TO ALL THE INTERFACES
  73. # JUST COMMENT OUT THE FOLLOWING LINE.
  74. # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  75. bind 127.0.0.1 -::1
  77. # Protected mode is a layer of security protection, in order to avoid that
  78. # Redis instances left open on the internet are accessed and exploited.
  79. #
  80. # When protected mode is on and if:
  81. #
  82. # 1) The server is not binding explicitly to a set of addresses using the
  83. #    "bind" directive.
  84. # 2) No password is configured.
  85. #
  86. # The server only accepts connections from clients connecting from the
  87. # IPv4 and IPv6 loopback addresses 127.0.0.1 and ::1, and from Unix domain
  88. # sockets.
  89. #
  90. # By default protected mode is enabled. You should disable it only if
  91. # you are sure you want clients from other hosts to connect to Redis
  92. # even if no authentication is configured, nor a specific set of interfaces
  93. # are explicitly listed using the "bind" directive.
  94. protected-mode yes
  96. # Accept connections on the specified port, default is 6379 (IANA #815344).
  97. # If port 0 is specified Redis will not listen on a TCP socket.
  98. port 6379
  100. # TCP listen() backlog.
  101. #
  102. # In high requests-per-second environments you need a high backlog in order
  103. # to avoid slow clients connection issues. Note that the Linux kernel
  104. # will silently truncate it to the value of /proc/sys/net/core/somaxconn so
  105. # make sure to raise both the value of somaxconn and tcp_max_syn_backlog
  106. # in order to get the desired effect.
  107. tcp-backlog 511
  109. # Unix socket.
  110. #
  111. # Specify the path for the Unix socket that will be used to listen for
  112. # incoming connections. There is no default, so Redis will not listen
  113. # on a unix socket when not specified.
  114. #
  115. # unixsocket /run/redis.sock
  116. # unixsocketperm 700
  118. # Close the connection after a client is idle for N seconds (0 to disable)
  119. timeout 0
  121. # TCP keepalive.
  122. #
  123. # If non-zero, use SO_KEEPALIVE to send TCP ACKs to clients in absence
  124. # of communication. This is useful for two reasons:
  125. #
  126. # 1) Detect dead peers.
  127. # 2) Force network equipment in the middle to consider the connection to be
  128. #    alive.
  129. #
  130. # On Linux, the specified value (in seconds) is the period used to send ACKs.
  131. # Note that to close the connection the double of the time is needed.
  132. # On other kernels the period depends on the kernel configuration.
  133. #
  134. # A reasonable value for this option is 300 seconds, which is the new
  135. # Redis default starting with Redis 3.2.1.
  136. tcp-keepalive 300
  138. ################################# TLS/SSL #####################################
  140. # By default, TLS/SSL is disabled. To enable it, the "tls-port" configuration
  141. # directive can be used to define TLS-listening ports. To enable TLS on the
  142. # default port, use:
  143. #
  144. # port 0
  145. # tls-port 6379
  147. # Configure a X.509 certificate and private key to use for authenticating the
  148. # server to connected clients, masters or cluster peers.  These files should be
  149. # PEM formatted.
  150. #
  151. # tls-cert-file redis.crt 
  152. # tls-key-file redis.key
  153. #
  154. # If the key file is encrypted using a passphrase, it can be included here
  155. # as well.
  156. #
  157. # tls-key-file-pass secret
  159. # Normally Redis uses the same certificate for both server functions (accepting
  160. # connections) and client functions (replicating from a master, establishing
  161. # cluster bus connections, etc.).
  162. #
  163. # Sometimes certificates are issued with attributes that designate them as
  164. # client-only or server-only certificates. In that case it may be desired to use
  165. # different certificates for incoming (server) and outgoing (client)
  166. # connections. To do that, use the following directives:
  167. #
  168. # tls-client-cert-file client.crt
  169. # tls-client-key-file client.key
  170. #
  171. # If the key file is encrypted using a passphrase, it can be included here
  172. # as well.
  173. #
  174. # tls-client-key-file-pass secret
  176. # Configure a DH parameters file to enable Diffie-Hellman (DH) key exchange:
  177. #
  178. # tls-dh-params-file redis.dh
  180. # Configure a CA certificate(s) bundle or directory to authenticate TLS/SSL
  181. # clients and peers.  Redis requires an explicit configuration of at least one
  182. # of these, and will not implicitly use the system wide configuration.
  183. #
  184. # tls-ca-cert-file ca.crt
  185. # tls-ca-cert-dir /etc/ssl/certs
  187. # By default, clients (including replica servers) on a TLS port are required
  188. # to authenticate using valid client side certificates.
  189. #
  190. # If "no" is specified, client certificates are not required and not accepted.
  191. # If "optional" is specified, client certificates are accepted and must be
  192. # valid if provided, but are not required.
  193. #
  194. # tls-auth-clients no
  195. # tls-auth-clients optional
  197. # By default, a Redis replica does not attempt to establish a TLS connection
  198. # with its master.
  199. #
  200. # Use the following directive to enable TLS on replication links.
  201. #
  202. # tls-replication yes
  204. # By default, the Redis Cluster bus uses a plain TCP connection. To enable
  205. # TLS for the bus protocol, use the following directive:
  206. #
  207. # tls-cluster yes
  209. # By default, only TLSv1.2 and TLSv1.3 are enabled and it is highly recommended
  210. # that older formally deprecated versions are kept disabled to reduce the attack surface.
  211. # You can explicitly specify TLS versions to support.
  212. # Allowed values are case insensitive and include "TLSv1", "TLSv1.1", "TLSv1.2",
  213. # "TLSv1.3" (OpenSSL >= 1.1.1) or any combination.
  214. # To enable only TLSv1.2 and TLSv1.3, use:
  215. #
  216. # tls-protocols "TLSv1.2 TLSv1.3"
  218. # Configure allowed ciphers.  See the ciphers(1ssl) manpage for more information
  219. # about the syntax of this string.
  220. #
  221. # Note: this configuration applies only to <= TLSv1.2.
  222. #
  223. # tls-ciphers DEFAULT:!MEDIUM
  225. # Configure allowed TLSv1.3 ciphersuites.  See the ciphers(1ssl) manpage for more
  226. # information about the syntax of this string, and specifically for TLSv1.3
  227. # ciphersuites.
  228. #
  229. # tls-ciphersuites TLS_CHACHA20_POLY1305_SHA256
  231. # When choosing a cipher, use the server's preference instead of the client
  232. # preference. By default, the server follows the client's preference.
  233. #
  234. # tls-prefer-server-ciphers yes
  236. # By default, TLS session caching is enabled to allow faster and less expensive
  237. # reconnections by clients that support it. Use the following directive to disable
  238. # caching.
  239. #
  240. # tls-session-caching no
  242. # Change the default number of TLS sessions cached. A zero value sets the cache
  243. # to unlimited size. The default size is 20480.
  244. #
  245. # tls-session-cache-size 5000
  247. # Change the default timeout of cached TLS sessions. The default timeout is 300
  248. # seconds.
  249. #
  250. # tls-session-cache-timeout 60
  252. ################################# GENERAL #####################################
  254. # By default Redis does not run as a daemon. Use 'yes' if you need it.
  255. # Note that Redis will write a pid file in /var/run/redis.pid when daemonized.
  256. # When Redis is supervised by upstart or systemd, this parameter has no impact.
  257. daemonize no
  259. # If you run Redis from upstart or systemd, Redis can interact with your
  260. # supervision tree. Options:
  261. #   supervised no      - no supervision interaction
  262. #   supervised upstart - signal upstart by putting Redis into SIGSTOP mode
  263. #                        requires "expect stop" in your upstart job config
  264. #   supervised systemd - signal systemd by writing READY=1 to $NOTIFY_SOCKET
  265. #                        on startup, and updating Redis status on a regular
  266. #                        basis.
  267. #   supervised auto    - detect upstart or systemd method based on
  268. #                        UPSTART_JOB or NOTIFY_SOCKET environment variables
  269. # Note: these supervision methods only signal "process is ready."
  270. #       They do not enable continuous pings back to your supervisor.
  271. #
  272. # The default is "no". To run under upstart/systemd, you can simply uncomment
  273. # the line below:
  274. #
  275. # supervised auto
  277. # If a pid file is specified, Redis writes it where specified at startup
  278. # and removes it at exit.
  279. #
  280. # When the server runs non daemonized, no pid file is created if none is
  281. # specified in the configuration. When the server is daemonized, the pid file
  282. # is used even if not specified, defaulting to "/var/run/redis.pid".
  283. #
  284. # Creating a pid file is best effort: if Redis is not able to create it
  285. # nothing bad happens, the server will start and run normally.
  286. #
  287. # Note that on modern Linux systems "/run/redis.pid" is more conforming
  288. # and should be used instead.
  289. pidfile /var/run/redis_6379.pid
  291. # Specify the server verbosity level.
  292. # This can be one of:
  293. # debug (a lot of information, useful for development/testing)
  294. # verbose (many rarely useful info, but not a mess like the debug level)
  295. # notice (moderately verbose, what you want in production probably)
  296. # warning (only very important / critical messages are logged)
  297. loglevel notice
  299. # Specify the log file name. Also the empty string can be used to force
  300. # Redis to log on the standard output. Note that if you use standard
  301. # output for logging but daemonize, logs will be sent to /dev/null
  302. logfile ""
  304. # To enable logging to the system logger, just set 'syslog-enabled' to yes,
  305. # and optionally update the other syslog parameters to suit your needs.
  306. # syslog-enabled no
  308. # Specify the syslog identity.
  309. # syslog-ident redis
  311. # Specify the syslog facility. Must be USER or between LOCAL0-LOCAL7.
  312. # syslog-facility local0
  314. # To disable the built in crash log, which will possibly produce cleaner core
  315. # dumps when they are needed, uncomment the following:
  316. #
  317. # crash-log-enabled no
  319. # To disable the fast memory check that's run as part of the crash log, which
  320. # will possibly let redis terminate sooner, uncomment the following:
  321. #
  322. # crash-memcheck-enabled no
  324. # Set the number of databases. The default database is DB 0, you can select
  325. # a different one on a per-connection basis using SELECT <dbid> where
  326. # dbid is a number between 0 and 'databases'-1
  327. databases 16
  329. # By default Redis shows an ASCII art logo only when started to log to the
  330. # standard output and if the standard output is a TTY and syslog logging is
  331. # disabled. Basically this means that normally a logo is displayed only in
  332. # interactive sessions.
  333. #
  334. # However it is possible to force the pre-4.0 behavior and always show a
  335. # ASCII art logo in startup logs by setting the following option to yes.
  336. always-show-logo no
  338. # By default, Redis modifies the process title (as seen in 'top' and 'ps') to
  339. # provide some runtime information. It is possible to disable this and leave
  340. # the process name as executed by setting the following to no.
  341. set-proc-title yes
  343. # When changing the process title, Redis uses the following template to construct
  344. # the modified title.
  345. #
  346. # Template variables are specified in curly brackets. The following variables are
  347. # supported:
  348. #
  349. # {title}           Name of process as executed if parent, or type of child process.
  350. # {listen-addr}     Bind address or '*' followed by TCP or TLS port listening on, or
  351. #                   Unix socket if only that's available.
  352. # {server-mode}     Special mode, i.e. "[sentinel]" or "[cluster]".
  353. # {port}            TCP port listening on, or 0.
  354. # {tls-port}        TLS port listening on, or 0.
  355. # {unixsocket}      Unix domain socket listening on, or "".
  356. # {config-file}     Name of configuration file used.
  357. #
  358. proc-title-template "{title} {listen-addr} {server-mode}"
  360. ################################ SNAPSHOTTING  ################################
  362. # Save the DB to disk.
  363. #
  364. # save <seconds> <changes>
  365. #
  366. # Redis will save the DB if both the given number of seconds and the given
  367. # number of write operations against the DB occurred.
  368. #
  369. # Snapshotting can be completely disabled with a single empty string argument
  370. # as in following example:
  371. #
  372. # save ""
  373. #
  374. # Unless specified otherwise, by default Redis will save the DB:
  375. #   * After 3600 seconds (an hour) if at least 1 key changed
  376. #   * After 300 seconds (5 minutes) if at least 100 keys changed
  377. #   * After 60 seconds if at least 10000 keys changed
  378. #
  379. # You can set these explicitly by uncommenting the three following lines.
  380. #
  381. # save 3600 1
  382. # save 300 100
  383. # save 60 10000
  385. # By default Redis will stop accepting writes if RDB snapshots are enabled
  386. # (at least one save point) and the latest background save failed.
  387. # This will make the user aware (in a hard way) that data is not persisting
  388. # on disk properly, otherwise chances are that no one will notice and some
  389. # disaster will happen.
  390. #
  391. # If the background saving process will start working again Redis will
  392. # automatically allow writes again.
  393. #
  394. # However if you have setup your proper monitoring of the Redis server
  395. # and persistence, you may want to disable this feature so that Redis will
  396. # continue to work as usual even if there are problems with disk,
  397. # permissions, and so forth.
  398. stop-writes-on-bgsave-error yes
  400. # Compress string objects using LZF when dump .rdb databases?
  401. # By default compression is enabled as it's almost always a win.
  402. # If you want to save some CPU in the saving child set it to 'no' but
  403. # the dataset will likely be bigger if you have compressible values or keys.
  404. rdbcompression yes
  406. # Since version 5 of RDB a CRC64 checksum is placed at the end of the file.
  407. # This makes the format more resistant to corruption but there is a performance
  408. # hit to pay (around 10%) when saving and loading RDB files, so you can disable it
  409. # for maximum performances.
  410. #
  411. # RDB files created with checksum disabled have a checksum of zero that will
  412. # tell the loading code to skip the check.
  413. rdbchecksum yes
  415. # Enables or disables full sanitation checks for ziplist and listpack etc when
  416. # loading an RDB or RESTORE payload. This reduces the chances of a assertion or
  417. # crash later on while processing commands.
  418. # Options:
  419. #   no         - Never perform full sanitation
  420. #   yes        - Always perform full sanitation
  421. #   clients    - Perform full sanitation only for user connections.
  422. #                Excludes: RDB files, RESTORE commands received from the master
  423. #                connection, and client connections which have the
  424. #                skip-sanitize-payload ACL flag.
  425. # The default should be 'clients' but since it currently affects cluster
  426. # resharding via MIGRATE, it is temporarily set to 'no' by default.
  427. #
  428. # sanitize-dump-payload no
  430. # The filename where to dump the DB
  431. dbfilename dump.rdb
  433. # Remove RDB files used by replication in instances without persistence
  434. # enabled. By default this option is disabled, however there are environments
  435. # where for regulations or other security concerns, RDB files persisted on
  436. # disk by masters in order to feed replicas, or stored on disk by replicas
  437. # in order to load them for the initial synchronization, should be deleted
  438. # ASAP. Note that this option ONLY WORKS in instances that have both AOF
  439. # and RDB persistence disabled, otherwise is completely ignored.
  440. #
  441. # An alternative (and sometimes better) way to obtain the same effect is
  442. # to use diskless replication on both master and replicas instances. However
  443. # in the case of replicas, diskless is not always an option.
  444. rdb-del-sync-files no
  446. # The working directory.
  447. #
  448. # The DB will be written inside this directory, with the filename specified
  449. # above using the 'dbfilename' configuration directive.
  450. #
  451. # The Append Only File will also be created inside this directory.
  452. #
  453. # Note that you must specify a directory here, not a file name.
  454. dir ./
  456. ################################# REPLICATION #################################
  458. # Master-Replica replication. Use replicaof to make a Redis instance a copy of
  459. # another Redis server. A few things to understand ASAP about Redis replication.
  460. #
  461. #   +------------------+      +---------------+
  462. #   |      Master      | ---> |    Replica    |
  463. #   | (receive writes) |      |  (exact copy) |
  464. #   +------------------+      +---------------+
  465. #
  466. # 1) Redis replication is asynchronous, but you can configure a master to
  467. #    stop accepting writes if it appears to be not connected with at least
  468. #    a given number of replicas.
  469. # 2) Redis replicas are able to perform a partial resynchronization with the
  470. #    master if the replication link is lost for a relatively small amount of
  471. #    time. You may want to configure the replication backlog size (see the next
  472. #    sections of this file) with a sensible value depending on your needs.
  473. # 3) Replication is automatic and does not need user intervention. After a
  474. #    network partition replicas automatically try to reconnect to masters
  475. #    and resynchronize with them.
  476. #
  477. # replicaof <masterip> <masterport>
  479. # If the master is password protected (using the "requirepass" configuration
  480. # directive below) it is possible to tell the replica to authenticate before
  481. # starting the replication synchronization process, otherwise the master will
  482. # refuse the replica request.
  483. #
  484. # masterauth <master-password>
  485. #
  486. # However this is not enough if you are using Redis ACLs (for Redis version
  487. # 6 or greater), and the default user is not capable of running the PSYNC
  488. # command and/or other commands needed for replication. In this case it's
  489. # better to configure a special user to use with replication, and specify the
  490. # masteruser configuration as such:
  491. #
  492. # masteruser <username>
  493. #
  494. # When masteruser is specified, the replica will authenticate against its
  495. # master using the new AUTH form: AUTH <username> <password>.
  497. # When a replica loses its connection with the master, or when the replication
  498. # is still in progress, the replica can act in two different ways:
  499. #
  500. # 1) if replica-serve-stale-data is set to 'yes' (the default) the replica will
  501. #    still reply to client requests, possibly with out of date data, or the
  502. #    data set may just be empty if this is the first synchronization.
  503. #
  504. # 2) If replica-serve-stale-data is set to 'no' the replica will reply with
  505. #    an error "SYNC with master in progress" to all commands except:
  506. #    INFO, REPLICAOF, AUTH, PING, SHUTDOWN, REPLCONF, ROLE, CONFIG, SUBSCRIBE,
  507. #    UNSUBSCRIBE, PSUBSCRIBE, PUNSUBSCRIBE, PUBLISH, PUBSUB, COMMAND, POST,
  508. #    HOST and LATENCY.
  509. #
  510. replica-serve-stale-data yes
  512. # You can configure a replica instance to accept writes or not. Writing against
  513. # a replica instance may be useful to store some ephemeral data (because data
  514. # written on a replica will be easily deleted after resync with the master) but
  515. # may also cause problems if clients are writing to it because of a
  516. # misconfiguration.
  517. #
  518. # Since Redis 2.6 by default replicas are read-only.
  519. #
  520. # Note: read only replicas are not designed to be exposed to untrusted clients
  521. # on the internet. It's just a protection layer against misuse of the instance.
  522. # Still a read only replica exports by default all the administrative commands
  523. # such as CONFIG, DEBUG, and so forth. To a limited extent you can improve
  524. # security of read only replicas using 'rename-command' to shadow all the
  525. # administrative / dangerous commands.
  526. replica-read-only yes
  528. # Replication SYNC strategy: disk or socket.
  529. #
  530. # New replicas and reconnecting replicas that are not able to continue the
  531. # replication process just receiving differences, need to do what is called a
  532. # "full synchronization". An RDB file is transmitted from the master to the
  533. # replicas.
  534. #
  535. # The transmission can happen in two different ways:
  536. #
  537. # 1) Disk-backed: The Redis master creates a new process that writes the RDB
  538. #                 file on disk. Later the file is transferred by the parent
  539. #                 process to the replicas incrementally.
  540. # 2) Diskless: The Redis master creates a new process that directly writes the
  541. #              RDB file to replica sockets, without touching the disk at all.
  542. #
  543. # With disk-backed replication, while the RDB file is generated, more replicas
  544. # can be queued and served with the RDB file as soon as the current child
  545. # producing the RDB file finishes its work. With diskless replication instead
  546. # once the transfer starts, new replicas arriving will be queued and a new
  547. # transfer will start when the current one terminates.
  548. #
  549. # When diskless replication is used, the master waits a configurable amount of
  550. # time (in seconds) before starting the transfer in the hope that multiple
  551. # replicas will arrive and the transfer can be parallelized.
  552. #
  553. # With slow disks and fast (large bandwidth) networks, diskless replication
  554. # works better.
  555. repl-diskless-sync no
  557. # When diskless replication is enabled, it is possible to configure the delay
  558. # the server waits in order to spawn the child that transfers the RDB via socket
  559. # to the replicas.
  560. #
  561. # This is important since once the transfer starts, it is not possible to serve
  562. # new replicas arriving, that will be queued for the next RDB transfer, so the
  563. # server waits a delay in order to let more replicas arrive.
  564. #
  565. # The delay is specified in seconds, and by default is 5 seconds. To disable
  566. # it entirely just set it to 0 seconds and the transfer will start ASAP.
  567. repl-diskless-sync-delay 5
  569. # -----------------------------------------------------------------------------
  570. # WARNING: RDB diskless load is experimental. Since in this setup the replica
  571. # does not immediately store an RDB on disk, it may cause data loss during
  572. # failovers. RDB diskless load + Redis modules not handling I/O reads may also
  573. # cause Redis to abort in case of I/O errors during the initial synchronization
  574. # stage with the master. Use only if you know what you are doing.
  575. # -----------------------------------------------------------------------------
  576. #
  577. # Replica can load the RDB it reads from the replication link directly from the
  578. # socket, or store the RDB to a file and read that file after it was completely
  579. # received from the master.
  580. #
  581. # In many cases the disk is slower than the network, and storing and loading
  582. # the RDB file may increase replication time (and even increase the master's
  583. # Copy on Write memory and salve buffers).
  584. # However, parsing the RDB file directly from the socket may mean that we have
  585. # to flush the contents of the current database before the full rdb was
  586. # received. For this reason we have the following options:
  587. #
  588. # "disabled"    - Don't use diskless load (store the rdb file to the disk first)
  589. # "on-empty-db" - Use diskless load only when it is completely safe.
  590. # "swapdb"      - Keep a copy of the current db contents in RAM while parsing
  591. #                 the data directly from the socket. note that this requires
  592. #                 sufficient memory, if you don't have it, you risk an OOM kill.
  593. repl-diskless-load disabled
  595. # Replicas send PINGs to server in a predefined interval. It's possible to
  596. # change this interval with the repl_ping_replica_period option. The default
  597. # value is 10 seconds.
  598. #
  599. # repl-ping-replica-period 10
  601. # The following option sets the replication timeout for:
  602. #
  603. # 1) Bulk transfer I/O during SYNC, from the point of view of replica.
  604. # 2) Master timeout from the point of view of replicas (data, pings).
  605. # 3) Replica timeout from the point of view of masters (REPLCONF ACK pings).
  606. #
  607. # It is important to make sure that this value is greater than the value
  608. # specified for repl-ping-replica-period otherwise a timeout will be detected
  609. # every time there is low traffic between the master and the replica. The default
  610. # value is 60 seconds.
  611. #
  612. # repl-timeout 60
  614. # Disable TCP_NODELAY on the replica socket after SYNC?
  615. #
  616. # If you select "yes" Redis will use a smaller number of TCP packets and
  617. # less bandwidth to send data to replicas. But this can add a delay for
  618. # the data to appear on the replica side, up to 40 milliseconds with
  619. # Linux kernels using a default configuration.
  620. #
  621. # If you select "no" the delay for data to appear on the replica side will
  622. # be reduced but more bandwidth will be used for replication.
  623. #
  624. # By default we optimize for low latency, but in very high traffic conditions
  625. # or when the master and replicas are many hops away, turning this to "yes" may
  626. # be a good idea.
  627. repl-disable-tcp-nodelay no
  629. # Set the replication backlog size. The backlog is a buffer that accumulates
  630. # replica data when replicas are disconnected for some time, so that when a
  631. # replica wants to reconnect again, often a full resync is not needed, but a
  632. # partial resync is enough, just passing the portion of data the replica
  633. # missed while disconnected.
  634. #
  635. # The bigger the replication backlog, the longer the replica can endure the
  636. # disconnect and later be able to perform a partial resynchronization.
  637. #
  638. # The backlog is only allocated if there is at least one replica connected.
  639. #
  640. # repl-backlog-size 1mb
  642. # After a master has no connected replicas for some time, the backlog will be
  643. # freed. The following option configures the amount of seconds that need to
  644. # elapse, starting from the time the last replica disconnected, for the backlog
  645. # buffer to be freed.
  646. #
  647. # Note that replicas never free the backlog for timeout, since they may be
  648. # promoted to masters later, and should be able to correctly "partially
  649. # resynchronize" with other replicas: hence they should always accumulate backlog.
  650. #
  651. # A value of 0 means to never release the backlog.
  652. #
  653. # repl-backlog-ttl 3600
  655. # The replica priority is an integer number published by Redis in the INFO
  656. # output. It is used by Redis Sentinel in order to select a replica to promote
  657. # into a master if the master is no longer working correctly.
  658. #
  659. # A replica with a low priority number is considered better for promotion, so
  660. # for instance if there are three replicas with priority 10, 100, 25 Sentinel
  661. # will pick the one with priority 10, that is the lowest.
  662. #
  663. # However a special priority of 0 marks the replica as not able to perform the
  664. # role of master, so a replica with priority of 0 will never be selected by
  665. # Redis Sentinel for promotion.
  666. #
  667. # By default the priority is 100.
  668. replica-priority 100
  670. # -----------------------------------------------------------------------------
  671. # By default, Redis Sentinel includes all replicas in its reports. A replica
  672. # can be excluded from Redis Sentinel's announcements. An unannounced replica
  673. # will be ignored by the 'sentinel replicas <master>' command and won't be
  674. # exposed to Redis Sentinel's clients.
  675. #
  676. # This option does not change the behavior of replica-priority. Even with
  677. # replica-announced set to 'no', the replica can be promoted to master. To
  678. # prevent this behavior, set replica-priority to 0.
  679. #
  680. # replica-announced yes
  682. # It is possible for a master to stop accepting writes if there are less than
  683. # N replicas connected, having a lag less or equal than M seconds.
  684. #
  685. # The N replicas need to be in "online" state.
  686. #
  687. # The lag in seconds, that must be <= the specified value, is calculated from
  688. # the last ping received from the replica, that is usually sent every second.
  689. #
  690. # This option does not GUARANTEE that N replicas will accept the write, but
  691. # will limit the window of exposure for lost writes in case not enough replicas
  692. # are available, to the specified number of seconds.
  693. #
  694. # For example to require at least 3 replicas with a lag <= 10 seconds use:
  695. #
  696. # min-replicas-to-write 3
  697. # min-replicas-max-lag 10
  698. #
  699. # Setting one or the other to 0 disables the feature.
  700. #
  701. # By default min-replicas-to-write is set to 0 (feature disabled) and
  702. # min-replicas-max-lag is set to 10.
  704. # A Redis master is able to list the address and port of the attached
  705. # replicas in different ways. For example the "INFO replication" section
  706. # offers this information, which is used, among other tools, by
  707. # Redis Sentinel in order to discover replica instances.
  708. # Another place where this info is available is in the output of the
  709. # "ROLE" command of a master.
  710. #
  711. # The listed IP address and port normally reported by a replica is
  712. # obtained in the following way:
  713. #
  714. #   IP: The address is auto detected by checking the peer address
  715. #   of the socket used by the replica to connect with the master.
  716. #
  717. #   Port: The port is communicated by the replica during the replication
  718. #   handshake, and is normally the port that the replica is using to
  719. #   listen for connections.
  720. #
  721. # However when port forwarding or Network Address Translation (NAT) is
  722. # used, the replica may actually be reachable via different IP and port
  723. # pairs. The following two options can be used by a replica in order to
  724. # report to its master a specific set of IP and port, so that both INFO
  725. # and ROLE will report those values.
  726. #
  727. # There is no need to use both the options if you need to override just
  728. # the port or the IP address.
  729. #
  730. # replica-announce-ip 5.5.5.5
  731. # replica-announce-port 1234
  733. ############################### KEYS TRACKING #################################
  735. # Redis implements server assisted support for client side caching of values.
  736. # This is implemented using an invalidation table that remembers, using
  737. # a radix key indexed by key name, what clients have which keys. In turn
  738. # this is used in order to send invalidation messages to clients. Please
  739. # check this page to understand more about the feature:
  740. #
  741. #   https://redis.io/topics/client-side-caching
  742. #
  743. # When tracking is enabled for a client, all the read only queries are assumed
  744. # to be cached: this will force Redis to store information in the invalidation
  745. # table. When keys are modified, such information is flushed away, and
  746. # invalidation messages are sent to the clients. However if the workload is
  747. # heavily dominated by reads, Redis could use more and more memory in order
  748. # to track the keys fetched by many clients.
  749. #
  750. # For this reason it is possible to configure a maximum fill value for the
  751. # invalidation table. By default it is set to 1M of keys, and once this limit
  752. # is reached, Redis will start to evict keys in the invalidation table
  753. # even if they were not modified, just to reclaim memory: this will in turn
  754. # force the clients to invalidate the cached values. Basically the table
  755. # maximum size is a trade off between the memory you want to spend server
  756. # side to track information about who cached what, and the ability of clients
  757. # to retain cached objects in memory.
  758. #
  759. # If you set the value to 0, it means there are no limits, and Redis will
  760. # retain as many keys as needed in the invalidation table.
  761. # In the "stats" INFO section, you can find information about the number of
  762. # keys in the invalidation table at every given moment.
  763. #
  764. # Note: when key tracking is used in broadcasting mode, no memory is used
  765. # in the server side so this setting is useless.
  766. #
  767. # tracking-table-max-keys 1000000
  769. ################################## SECURITY ###################################
  771. # Warning: since Redis is pretty fast, an outside user can try up to
  772. # 1 million passwords per second against a modern box. This means that you
  773. # should use very strong passwords, otherwise they will be very easy to break.
  774. # Note that because the password is really a shared secret between the client
  775. # and the server, and should not be memorized by any human, the password
  776. # can be easily a long string from /dev/urandom or whatever, so by using a
  777. # long and unguessable password no brute force attack will be possible.
  779. # Redis ACL users are defined in the following format:
  780. #
  781. #   user <username> ... acl rules ...
  782. #
  783. # For example:
  784. #
  785. #   user worker +@list +@connection ~jobs:* on >ffa9203c493aa99
  786. #
  787. # The special username "default" is used for new connections. If this user
  788. # has the "nopass" rule, then new connections will be immediately authenticated
  789. # as the "default" user without the need of any password provided via the
  790. # AUTH command. Otherwise if the "default" user is not flagged with "nopass"
  791. # the connections will start in not authenticated state, and will require
  792. # AUTH (or the HELLO command AUTH option) in order to be authenticated and
  793. # start to work.
  794. #
  795. # The ACL rules that describe what a user can do are the following:
  796. #
  797. #  on           Enable the user: it is possible to authenticate as this user.
  798. #  off          Disable the user: it's no longer possible to authenticate
  799. #               with this user, however the already authenticated connections
  800. #               will still work.
  801. #  skip-sanitize-payload    RESTORE dump-payload sanitation is skipped.
  802. #  sanitize-payload         RESTORE dump-payload is sanitized (default).
  803. #  +<command>   Allow the execution of that command
  804. #  -<command>   Disallow the execution of that command
  805. #  +@<category> Allow the execution of all the commands in such category
  806. #               with valid categories are like @admin, @set, @sortedset, ...
  807. #               and so forth, see the full list in the server.c file where
  808. #               the Redis command table is described and defined.
  809. #               The special category @all means all the commands, but currently
  810. #               present in the server, and that will be loaded in the future
  811. #               via modules.
  812. #  +<command>|subcommand    Allow a specific subcommand of an otherwise
  813. #                           disabled command. Note that this form is not
  814. #                           allowed as negative like -DEBUG|SEGFAULT, but
  815. #                           only additive starting with "+".
  816. #  allcommands  Alias for +@all. Note that it implies the ability to execute
  817. #               all the future commands loaded via the modules system.
  818. #  nocommands   Alias for -@all.
  819. #  ~<pattern>   Add a pattern of keys that can be mentioned as part of
  820. #               commands. For instance ~* allows all the keys. The pattern
  821. #               is a glob-style pattern like the one of KEYS.
  822. #               It is possible to specify multiple patterns.
  823. #  allkeys      Alias for ~*
  824. #  resetkeys    Flush the list of allowed keys patterns.
  825. #  &<pattern>   Add a glob-style pattern of Pub/Sub channels that can be
  826. #               accessed by the user. It is possible to specify multiple channel
  827. #               patterns.
  828. #  allchannels  Alias for &*
  829. #  resetchannels            Flush the list of allowed channel patterns.
  830. #  ><password>  Add this password to the list of valid password for the user.
  831. #               For example >mypass will add "mypass" to the list.
  832. #               This directive clears the "nopass" flag (see later).
  833. #  <<password>  Remove this password from the list of valid passwords.
  834. #  nopass       All the set passwords of the user are removed, and the user
  835. #               is flagged as requiring no password: it means that every
  836. #               password will work against this user. If this directive is
  837. #               used for the default user, every new connection will be
  838. #               immediately authenticated with the default user without
  839. #               any explicit AUTH command required. Note that the "resetpass"
  840. #               directive will clear this condition.
  841. #  resetpass    Flush the list of allowed passwords. Moreover removes the
  842. #               "nopass" status. After "resetpass" the user has no associated
  843. #               passwords and there is no way to authenticate without adding
  844. #               some password (or setting it as "nopass" later).
  845. #  reset        Performs the following actions: resetpass, resetkeys, off,
  846. #               -@all. The user returns to the same state it has immediately
  847. #               after its creation.
  848. #
  849. # ACL rules can be specified in any order: for instance you can start with
  850. # passwords, then flags, or key patterns. However note that the additive
  851. # and subtractive rules will CHANGE MEANING depending on the ordering.
  852. # For instance see the following example:
  853. #
  854. #   user alice on +@all -DEBUG ~* >somepassword
  855. #
  856. # This will allow "alice" to use all the commands with the exception of the
  857. # DEBUG command, since +@all added all the commands to the set of the commands
  858. # alice can use, and later DEBUG was removed. However if we invert the order
  859. # of two ACL rules the result will be different:
  860. #
  861. #   user alice on -DEBUG +@all ~* >somepassword
  862. #
  863. # Now DEBUG was removed when alice had yet no commands in the set of allowed
  864. # commands, later all the commands are added, so the user will be able to
  865. # execute everything.
  866. #
  867. # Basically ACL rules are processed left-to-right.
  868. #
  869. # For more information about ACL configuration please refer to
  870. # the Redis web site at https://redis.io/topics/acl
  872. # ACL LOG
  873. #
  874. # The ACL Log tracks failed commands and authentication events associated
  875. # with ACLs. The ACL Log is useful to troubleshoot failed commands blocked 
  876. # by ACLs. The ACL Log is stored in memory. You can reclaim memory with 
  877. # ACL LOG RESET. Define the maximum entry length of the ACL Log below.
  878. acllog-max-len 128
  880. # Using an external ACL file
  881. #
  882. # Instead of configuring users here in this file, it is possible to use
  883. # a stand-alone file just listing users. The two methods cannot be mixed:
  884. # if you configure users here and at the same time you activate the external
  885. # ACL file, the server will refuse to start.
  886. #
  887. # The format of the external ACL user file is exactly the same as the
  888. # format that is used inside redis.conf to describe users.
  889. #
  890. # aclfile /etc/redis/users.acl
  892. # IMPORTANT NOTE: starting with Redis 6 "requirepass" is just a compatibility
  893. # layer on top of the new ACL system. The option effect will be just setting
  894. # the password for the default user. Clients will still authenticate using
  895. # AUTH <password> as usually, or more explicitly with AUTH default <password>
  896. # if they follow the new protocol: both will work.
  897. #
  898. # The requirepass is not compatable with aclfile option and the ACL LOAD
  899. # command, these will cause requirepass to be ignored.
  900. #
  901. # requirepass foobared
  903. # New users are initialized with restrictive permissions by default, via the
  904. # equivalent of this ACL rule 'off resetkeys -@all'. Starting with Redis 6.2, it
  905. # is possible to manage access to Pub/Sub channels with ACL rules as well. The
  906. # default Pub/Sub channels permission if new users is controlled by the 
  907. # acl-pubsub-default configuration directive, which accepts one of these values:
  908. #
  909. # allchannels: grants access to all Pub/Sub channels
  910. # resetchannels: revokes access to all Pub/Sub channels
  911. #
  912. # To ensure backward compatibility while upgrading Redis 6.0, acl-pubsub-default
  913. # defaults to the 'allchannels' permission.
  914. #
  915. # Future compatibility note: it is very likely that in a future version of Redis
  916. # the directive's default of 'allchannels' will be changed to 'resetchannels' in
  917. # order to provide better out-of-the-box Pub/Sub security. Therefore, it is
  918. # recommended that you explicitly define Pub/Sub permissions for all users
  919. # rather then rely on implicit default values. Once you've set explicit
  920. # Pub/Sub for all existing users, you should uncomment the following line.
  921. #
  922. # acl-pubsub-default resetchannels
  924. # Command renaming (DEPRECATED).
  925. #
  926. # ------------------------------------------------------------------------
  927. # WARNING: avoid using this option if possible. Instead use ACLs to remove
  928. # commands from the default user, and put them only in some admin user you
  929. # create for administrative purposes.
  930. # ------------------------------------------------------------------------
  931. #
  932. # It is possible to change the name of dangerous commands in a shared
  933. # environment. For instance the CONFIG command may be renamed into something
  934. # hard to guess so that it will still be available for internal-use tools
  935. # but not available for general clients.
  936. #
  937. # Example:
  938. #
  939. # rename-command CONFIG b840fc02d524045429941cc15f59e41cb7be6c52
  940. #
  941. # It is also possible to completely kill a command by renaming it into
  942. # an empty string:
  943. #
  944. # rename-command CONFIG ""
  945. #
  946. # Please note that changing the name of commands that are logged into the
  947. # AOF file or transmitted to replicas may cause problems.
  949. ################################### CLIENTS ####################################
  951. # Set the max number of connected clients at the same time. By default
  952. # this limit is set to 10000 clients, however if the Redis server is not
  953. # able to configure the process file limit to allow for the specified limit
  954. # the max number of allowed clients is set to the current file limit
  955. # minus 32 (as Redis reserves a few file descriptors for internal uses).
  956. #
  957. # Once the limit is reached Redis will close all the new connections sending
  958. # an error 'max number of clients reached'.
  959. #
  960. # IMPORTANT: When Redis Cluster is used, the max number of connections is also
  961. # shared with the cluster bus: every node in the cluster will use two
  962. # connections, one incoming and another outgoing. It is important to size the
  963. # limit accordingly in case of very large clusters.
  964. #
  965. # maxclients 10000
  967. ############################## MEMORY MANAGEMENT ################################
  969. # Set a memory usage limit to the specified amount of bytes.
  970. # When the memory limit is reached Redis will try to remove keys
  971. # according to the eviction policy selected (see maxmemory-policy).
  972. #
  973. # If Redis can't remove keys according to the policy, or if the policy is
  974. # set to 'noeviction', Redis will start to reply with errors to commands
  975. # that would use more memory, like SET, LPUSH, and so on, and will continue
  976. # to reply to read-only commands like GET.
  977. #
  978. # This option is usually useful when using Redis as an LRU or LFU cache, or to
  979. # set a hard memory limit for an instance (using the 'noeviction' policy).
  980. #
  981. # WARNING: If you have replicas attached to an instance with maxmemory on,
  982. # the size of the output buffers needed to feed the replicas are subtracted
  983. # from the used memory count, so that network problems / resyncs will
  984. # not trigger a loop where keys are evicted, and in turn the output
  985. # buffer of replicas is full with DELs of keys evicted triggering the deletion
  986. # of more keys, and so forth until the database is completely emptied.
  987. #
  988. # In short... if you have replicas attached it is suggested that you set a lower
  989. # limit for maxmemory so that there is some free RAM on the system for replica
  990. # output buffers (but this is not needed if the policy is 'noeviction').
  991. #
  992. # maxmemory <bytes>
  994. # MAXMEMORY POLICY: how Redis will select what to remove when maxmemory
  995. # is reached. You can select one from the following behaviors:
  996. #
  997. # volatile-lru -> Evict using approximated LRU, only keys with an expire set.
  998. # allkeys-lru -> Evict any key using approximated LRU.
  999. # volatile-lfu -> Evict using approximated LFU, only keys with an expire set.
  1000. # allkeys-lfu -> Evict any key using approximated LFU.
  1001. # volatile-random -> Remove a random key having an expire set.
  1002. # allkeys-random -> Remove a random key, any key.
  1003. # volatile-ttl -> Remove the key with the nearest expire time (minor TTL)
  1004. # noeviction -> Don't evict anything, just return an error on write operations.
  1005. #
  1006. # LRU means Least Recently Used
  1007. # LFU means Least Frequently Used
  1008. #
  1009. # Both LRU, LFU and volatile-ttl are implemented using approximated
  1010. # randomized algorithms.
  1011. #
  1012. # Note: with any of the above policies, when there are no suitable keys for
  1013. # eviction, Redis will return an error on write operations that require
  1014. # more memory. These are usually commands that create new keys, add data or
  1015. # modify existing keys. A few examples are: SET, INCR, HSET, LPUSH, SUNIONSTORE,
  1016. # SORT (due to the STORE argument), and EXEC (if the transaction includes any
  1017. # command that requires memory).
  1018. #
  1019. # The default is:
  1020. #
  1021. # maxmemory-policy noeviction
  1023. # LRU, LFU and minimal TTL algorithms are not precise algorithms but approximated
  1024. # algorithms (in order to save memory), so you can tune it for speed or
  1025. # accuracy. By default Redis will check five keys and pick the one that was
  1026. # used least recently, you can change the sample size using the following
  1027. # configuration directive.
  1028. #
  1029. # The default of 5 produces good enough results. 10 Approximates very closely
  1030. # true LRU but costs more CPU. 3 is faster but not very accurate.
  1031. #
  1032. # maxmemory-samples 5
  1034. # Eviction processing is designed to function well with the default setting.
  1035. # If there is an unusually large amount of write traffic, this value may need to
  1036. # be increased.  Decreasing this value may reduce latency at the risk of 
  1037. # eviction processing effectiveness
  1038. #   0 = minimum latency, 10 = default, 100 = process without regard to latency
  1039. #
  1040. # maxmemory-eviction-tenacity 10
  1042. # Starting from Redis 5, by default a replica will ignore its maxmemory setting
  1043. # (unless it is promoted to master after a failover or manually). It means
  1044. # that the eviction of keys will be just handled by the master, sending the
  1045. # DEL commands to the replica as keys evict in the master side.
  1046. #
  1047. # This behavior ensures that masters and replicas stay consistent, and is usually
  1048. # what you want, however if your replica is writable, or you want the replica
  1049. # to have a different memory setting, and you are sure all the writes performed
  1050. # to the replica are idempotent, then you may change this default (but be sure
  1051. # to understand what you are doing).
  1052. #
  1053. # Note that since the replica by default does not evict, it may end using more
  1054. # memory than the one set via maxmemory (there are certain buffers that may
  1055. # be larger on the replica, or data structures may sometimes take more memory
  1056. # and so forth). So make sure you monitor your replicas and make sure they
  1057. # have enough memory to never hit a real out-of-memory condition before the
  1058. # master hits the configured maxmemory setting.
  1059. #
  1060. # replica-ignore-maxmemory yes
  1062. # Redis reclaims expired keys in two ways: upon access when those keys are
  1063. # found to be expired, and also in background, in what is called the
  1064. # "active expire key". The key space is slowly and interactively scanned
  1065. # looking for expired keys to reclaim, so that it is possible to free memory
  1066. # of keys that are expired and will never be accessed again in a short time.
  1067. #
  1068. # The default effort of the expire cycle will try to avoid having more than
  1069. # ten percent of expired keys still in memory, and will try to avoid consuming
  1070. # more than 25% of total memory and to add latency to the system. However
  1071. # it is possible to increase the expire "effort" that is normally set to
  1072. # "1", to a greater value, up to the value "10". At its maximum value the
  1073. # system will use more CPU, longer cycles (and technically may introduce
  1074. # more latency), and will tolerate less already expired keys still present
  1075. # in the system. It's a tradeoff between memory, CPU and latency.
  1076. #
  1077. # active-expire-effort 1
  1079. ############################# LAZY FREEING ####################################
  1081. # Redis has two primitives to delete keys. One is called DEL and is a blocking
  1082. # deletion of the object. It means that the server stops processing new commands
  1083. # in order to reclaim all the memory associated with an object in a synchronous
  1084. # way. If the key deleted is associated with a small object, the time needed
  1085. # in order to execute the DEL command is very small and comparable to most other
  1086. # O(1) or O(log_N) commands in Redis. However if the key is associated with an
  1087. # aggregated value containing millions of elements, the server can block for
  1088. # a long time (even seconds) in order to complete the operation.
  1089. #
  1090. # For the above reasons Redis also offers non blocking deletion primitives
  1091. # such as UNLINK (non blocking DEL) and the ASYNC option of FLUSHALL and
  1092. # FLUSHDB commands, in order to reclaim memory in background. Those commands
  1093. # are executed in constant time. Another thread will incrementally free the
  1094. # object in the background as fast as possible.
  1095. #
  1096. # DEL, UNLINK and ASYNC option of FLUSHALL and FLUSHDB are user-controlled.
  1097. # It's up to the design of the application to understand when it is a good
  1098. # idea to use one or the other. However the Redis server sometimes has to
  1099. # delete keys or flush the whole database as a side effect of other operations.
  1100. # Specifically Redis deletes objects independently of a user call in the
  1101. # following scenarios:
  1102. #
  1103. # 1) On eviction, because of the maxmemory and maxmemory policy configurations,
  1104. #    in order to make room for new data, without going over the specified
  1105. #    memory limit.
  1106. # 2) Because of expire: when a key with an associated time to live (see the
  1107. #    EXPIRE command) must be deleted from memory.
  1108. # 3) Because of a side effect of a command that stores data on a key that may
  1109. #    already exist. For example the RENAME command may delete the old key
  1110. #    content when it is replaced with another one. Similarly SUNIONSTORE
  1111. #    or SORT with STORE option may delete existing keys. The SET command
  1112. #    itself removes any old content of the specified key in order to replace
  1113. #    it with the specified string.
  1114. # 4) During replication, when a replica performs a full resynchronization with
  1115. #    its master, the content of the whole database is removed in order to
  1116. #    load the RDB file just transferred.
  1117. #
  1118. # In all the above cases the default is to delete objects in a blocking way,
  1119. # like if DEL was called. However you can configure each case specifically
  1120. # in order to instead release memory in a non-blocking way like if UNLINK
  1121. # was called, using the following configuration directives.
  1123. lazyfree-lazy-eviction no
  1124. lazyfree-lazy-expire no
  1125. lazyfree-lazy-server-del no
  1126. replica-lazy-flush no
  1128. # It is also possible, for the case when to replace the user code DEL calls
  1129. # with UNLINK calls is not easy, to modify the default behavior of the DEL
  1130. # command to act exactly like UNLINK, using the following configuration
  1131. # directive:
  1133. lazyfree-lazy-user-del no
  1135. # FLUSHDB, FLUSHALL, and SCRIPT FLUSH support both asynchronous and synchronous
  1136. # deletion, which can be controlled by passing the [SYNC|ASYNC] flags into the
  1137. # commands. When neither flag is passed, this directive will be used to determine
  1138. # if the data should be deleted asynchronously.
  1140. lazyfree-lazy-user-flush no
  1142. ################################ THREADED I/O #################################
  1144. # Redis is mostly single threaded, however there are certain threaded
  1145. # operations such as UNLINK, slow I/O accesses and other things that are
  1146. # performed on side threads.
  1147. #
  1148. # Now it is also possible to handle Redis clients socket reads and writes
  1149. # in different I/O threads. Since especially writing is so slow, normally
  1150. # Redis users use pipelining in order to speed up the Redis performances per
  1151. # core, and spawn multiple instances in order to scale more. Using I/O
  1152. # threads it is possible to easily speedup two times Redis without resorting
  1153. # to pipelining nor sharding of the instance.
  1154. #
  1155. # By default threading is disabled, we suggest enabling it only in machines
  1156. # that have at least 4 or more cores, leaving at least one spare core.
  1157. # Using more than 8 threads is unlikely to help much. We also recommend using
  1158. # threaded I/O only if you actually have performance problems, with Redis
  1159. # instances being able to use a quite big percentage of CPU time, otherwise
  1160. # there is no point in using this feature.
  1161. #
  1162. # So for instance if you have a four cores boxes, try to use 2 or 3 I/O
  1163. # threads, if you have a 8 cores, try to use 6 threads. In order to
  1164. # enable I/O threads use the following configuration directive:
  1165. #
  1166. # io-threads 4
  1167. #
  1168. # Setting io-threads to 1 will just use the main thread as usual.
  1169. # When I/O threads are enabled, we only use threads for writes, that is
  1170. # to thread the write(2) syscall and transfer the client buffers to the
  1171. # socket. However it is also possible to enable threading of reads and
  1172. # protocol parsing using the following configuration directive, by setting
  1173. # it to yes:
  1174. #
  1175. # io-threads-do-reads no
  1176. #
  1177. # Usually threading reads doesn't help much.
  1178. #
  1179. # NOTE 1: This configuration directive cannot be changed at runtime via
  1180. # CONFIG SET. Aso this feature currently does not work when SSL is
  1181. # enabled.
  1182. #
  1183. # NOTE 2: If you want to test the Redis speedup using redis-benchmark, make
  1184. # sure you also run the benchmark itself in threaded mode, using the
  1185. # --threads option to match the number of Redis threads, otherwise you'll not
  1186. # be able to notice the improvements.
  1188. ############################ KERNEL OOM CONTROL ##############################
  1190. # On Linux, it is possible to hint the kernel OOM killer on what processes
  1191. # should be killed first when out of memory.
  1192. #
  1193. # Enabling this feature makes Redis actively control the oom_score_adj value
  1194. # for all its processes, depending on their role. The default scores will
  1195. # attempt to have background child processes killed before all others, and
  1196. # replicas killed before masters.
  1197. #
  1198. # Redis supports three options:
  1199. #
  1200. # no:       Don't make changes to oom-score-adj (default).
  1201. # yes:      Alias to "relative" see below.
  1202. # absolute: Values in oom-score-adj-values are written as is to the kernel.
  1203. # relative: Values are used relative to the initial value of oom_score_adj when
  1204. #           the server starts and are then clamped to a range of -1000 to 1000.
  1205. #           Because typically the initial value is 0, they will often match the
  1206. #           absolute values.
  1207. oom-score-adj no
  1209. # When oom-score-adj is used, this directive controls the specific values used
  1210. # for master, replica and background child processes. Values range -2000 to
  1211. # 2000 (higher means more likely to be killed).
  1212. #
  1213. # Unprivileged processes (not root, and without CAP_SYS_RESOURCE capabilities)
  1214. # can freely increase their value, but not decrease it below its initial
  1215. # settings. This means that setting oom-score-adj to "relative" and setting the
  1216. # oom-score-adj-values to positive values will always succeed.
  1217. oom-score-adj-values 0 200 800
  1220. #################### KERNEL transparent hugepage CONTROL ######################
  1222. # Usually the kernel Transparent Huge Pages control is set to "madvise" or
  1223. # or "never" by default (/sys/kernel/mm/transparent_hugepage/enabled), in which
  1224. # case this config has no effect. On systems in which it is set to "always",
  1225. # redis will attempt to disable it specifically for the redis process in order
  1226. # to avoid latency problems specifically with fork(2) and CoW.
  1227. # If for some reason you prefer to keep it enabled, you can set this config to
  1228. # "no" and the kernel global to "always".
  1230. disable-thp yes
  1232. ############################## APPEND ONLY MODE ###############################
  1234. # By default Redis asynchronously dumps the dataset on disk. This mode is
  1235. # good enough in many applications, but an issue with the Redis process or
  1236. # a power outage may result into a few minutes of writes lost (depending on
  1237. # the configured save points).
  1238. #
  1239. # The Append Only File is an alternative persistence mode that provides
  1240. # much better durability. For instance using the default data fsync policy
  1241. # (see later in the config file) Redis can lose just one second of writes in a
  1242. # dramatic event like a server power outage, or a single write if something
  1243. # wrong with the Redis process itself happens, but the operating system is
  1244. # still running correctly.
  1245. #
  1246. # AOF and RDB persistence can be enabled at the same time without problems.
  1247. # If the AOF is enabled on startup Redis will load the AOF, that is the file
  1248. # with the better durability guarantees.
  1249. #
  1250. # Please check https://redis.io/topics/persistence for more information.
  1252. appendonly no
  1254. # The name of the append only file (default: "appendonly.aof")
  1256. appendfilename "appendonly.aof"
  1258. # The fsync() call tells the Operating System to actually write data on disk
  1259. # instead of waiting for more data in the output buffer. Some OS will really flush
  1260. # data on disk, some other OS will just try to do it ASAP.
  1261. #
  1262. # Redis supports three different modes:
  1263. #
  1264. # no: don't fsync, just let the OS flush the data when it wants. Faster.
  1265. # always: fsync after every write to the append only log. Slow, Safest.
  1266. # everysec: fsync only one time every second. Compromise.
  1267. #
  1268. # The default is "everysec", as that's usually the right compromise between
  1269. # speed and data safety. It's up to you to understand if you can relax this to
  1270. # "no" that will let the operating system flush the output buffer when
  1271. # it wants, for better performances (but if you can live with the idea of
  1272. # some data loss consider the default persistence mode that's snapshotting),
  1273. # or on the contrary, use "always" that's very slow but a bit safer than
  1274. # everysec.
  1275. #
  1276. # More details please check the following article:
  1277. # http://antirez.com/post/redis-persistence-demystified.html
  1278. #
  1279. # If unsure, use "everysec".
  1281. # appendfsync always
  1282. appendfsync everysec
  1283. # appendfsync no
  1285. # When the AOF fsync policy is set to always or everysec, and a background
  1286. # saving process (a background save or AOF log background rewriting) is
  1287. # performing a lot of I/O against the disk, in some Linux configurations
  1288. # Redis may block too long on the fsync() call. Note that there is no fix for
  1289. # this currently, as even performing fsync in a different thread will block
  1290. # our synchronous write(2) call.
  1291. #
  1292. # In order to mitigate this problem it's possible to use the following option
  1293. # that will prevent fsync() from being called in the main process while a
  1294. # BGSAVE or BGREWRITEAOF is in progress.
  1295. #
  1296. # This means that while another child is saving, the durability of Redis is
  1297. # the same as "appendfsync none". In practical terms, this means that it is
  1298. # possible to lose up to 30 seconds of log in the worst scenario (with the
  1299. # default Linux settings).
  1300. #
  1301. # If you have latency problems turn this to "yes". Otherwise leave it as
  1302. # "no" that is the safest pick from the point of view of durability.
  1304. no-appendfsync-on-rewrite no
  1306. # Automatic rewrite of the append only file.
  1307. # Redis is able to automatically rewrite the log file implicitly calling
  1308. # BGREWRITEAOF when the AOF log size grows by the specified percentage.
  1309. #
  1310. # This is how it works: Redis remembers the size of the AOF file after the
  1311. # latest rewrite (if no rewrite has happened since the restart, the size of
  1312. # the AOF at startup is used).
  1313. #
  1314. # This base size is compared to the current size. If the current size is
  1315. # bigger than the specified percentage, the rewrite is triggered. Also
  1316. # you need to specify a minimal size for the AOF file to be rewritten, this
  1317. # is useful to avoid rewriting the AOF file even if the percentage increase
  1318. # is reached but it is still pretty small.
  1319. #
  1320. # Specify a percentage of zero in order to disable the automatic AOF
  1321. # rewrite feature.
  1323. auto-aof-rewrite-percentage 100
  1324. auto-aof-rewrite-min-size 64mb
  1326. # An AOF file may be found to be truncated at the end during the Redis
  1327. # startup process, when the AOF data gets loaded back into memory.
  1328. # This may happen when the system where Redis is running
  1329. # crashes, especially when an ext4 filesystem is mounted without the
  1330. # data=ordered option (however this can't happen when Redis itself
  1331. # crashes or aborts but the operating system still works correctly).
  1332. #
  1333. # Redis can either exit with an error when this happens, or load as much
  1334. # data as possible (the default now) and start if the AOF file is found
  1335. # to be truncated at the end. The following option controls this behavior.
  1336. #
  1337. # If aof-load-truncated is set to yes, a truncated AOF file is loaded and
  1338. # the Redis server starts emitting a log to inform the user of the event.
  1339. # Otherwise if the option is set to no, the server aborts with an error
  1340. # and refuses to start. When the option is set to no, the user requires
  1341. # to fix the AOF file using the "redis-check-aof" utility before to restart
  1342. # the server.
  1343. #
  1344. # Note that if the AOF file will be found to be corrupted in the middle
  1345. # the server will still exit with an error. This option only applies when
  1346. # Redis will try to read more data from the AOF file but not enough bytes
  1347. # will be found.
  1348. aof-load-truncated yes
  1350. # When rewriting the AOF file, Redis is able to use an RDB preamble in the
  1351. # AOF file for faster rewrites and recoveries. When this option is turned
  1352. # on the rewritten AOF file is composed of two different stanzas:
  1353. #
  1354. #   [RDB file][AOF tail]
  1355. #
  1356. # When loading, Redis recognizes that the AOF file starts with the "REDIS"
  1357. # string and loads the prefixed RDB file, then continues loading the AOF
  1358. # tail.
  1359. aof-use-rdb-preamble yes
  1361. ################################ LUA SCRIPTING  ###############################
  1363. # Max execution time of a Lua script in milliseconds.
  1364. #
  1365. # If the maximum execution time is reached Redis will log that a script is
  1366. # still in execution after the maximum allowed time and will start to
  1367. # reply to queries with an error.
  1368. #
  1369. # When a long running script exceeds the maximum execution time only the
  1370. # SCRIPT KILL and SHUTDOWN NOSAVE commands are available. The first can be
  1371. # used to stop a script that did not yet call any write commands. The second
  1372. # is the only way to shut down the server in the case a write command was
  1373. # already issued by the script but the user doesn't want to wait for the natural
  1374. # termination of the script.
  1375. #
  1376. # Set it to 0 or a negative value for unlimited execution without warnings.
  1377. lua-time-limit 5000
  1379. ################################ REDIS CLUSTER  ###############################
  1381. # Normal Redis instances can't be part of a Redis Cluster; only nodes that are
  1382. # started as cluster nodes can. In order to start a Redis instance as a
  1383. # cluster node enable the cluster support uncommenting the following:
  1384. #
  1385. # cluster-enabled yes
  1387. # Every cluster node has a cluster configuration file. This file is not
  1388. # intended to be edited by hand. It is created and updated by Redis nodes.
  1389. # Every Redis Cluster node requires a different cluster configuration file.
  1390. # Make sure that instances running in the same system do not have
  1391. # overlapping cluster configuration file names.
  1392. #
  1393. # cluster-config-file nodes-6379.conf
  1395. # Cluster node timeout is the amount of milliseconds a node must be unreachable
  1396. # for it to be considered in failure state.
  1397. # Most other internal time limits are a multiple of the node timeout.
  1398. #
  1399. # cluster-node-timeout 15000
  1401. # A replica of a failing master will avoid to start a failover if its data
  1402. # looks too old.
  1403. #
  1404. # There is no simple way for a replica to actually have an exact measure of
  1405. # its "data age", so the following two checks are performed:
  1406. #
  1407. # 1) If there are multiple replicas able to failover, they exchange messages
  1408. #    in order to try to give an advantage to the replica with the best
  1409. #    replication offset (more data from the master processed).
  1410. #    Replicas will try to get their rank by offset, and apply to the start
  1411. #    of the failover a delay proportional to their rank.
  1412. #
  1413. # 2) Every single replica computes the time of the last interaction with
  1414. #    its master. This can be the last ping or command received (if the master
  1415. #    is still in the "connected" state), or the time that elapsed since the
  1416. #    disconnection with the master (if the replication link is currently down).
  1417. #    If the last interaction is too old, the replica will not try to failover
  1418. #    at all.
  1419. #
  1420. # The point "2" can be tuned by user. Specifically a replica will not perform
  1421. # the failover if, since the last interaction with the master, the time
  1422. # elapsed is greater than:
  1423. #
  1424. #   (node-timeout * cluster-replica-validity-factor) + repl-ping-replica-period
  1425. #
  1426. # So for example if node-timeout is 30 seconds, and the cluster-replica-validity-factor
  1427. # is 10, and assuming a default repl-ping-replica-period of 10 seconds, the
  1428. # replica will not try to failover if it was not able to talk with the master
  1429. # for longer than 310 seconds.
  1430. #
  1431. # A large cluster-replica-validity-factor may allow replicas with too old data to failover
  1432. # a master, while a too small value may prevent the cluster from being able to
  1433. # elect a replica at all.
  1434. #
  1435. # For maximum availability, it is possible to set the cluster-replica-validity-factor
  1436. # to a value of 0, which means, that replicas will always try to failover the
  1437. # master regardless of the last time they interacted with the master.
  1438. # (However they'll always try to apply a delay proportional to their
  1439. # offset rank).
  1440. #
  1441. # Zero is the only value able to guarantee that when all the partitions heal
  1442. # the cluster will always be able to continue.
  1443. #
  1444. # cluster-replica-validity-factor 10
  1446. # Cluster replicas are able to migrate to orphaned masters, that are masters
  1447. # that are left without working replicas. This improves the cluster ability
  1448. # to resist to failures as otherwise an orphaned master can't be failed over
  1449. # in case of failure if it has no working replicas.
  1450. #
  1451. # Replicas migrate to orphaned masters only if there are still at least a
  1452. # given number of other working replicas for their old master. This number
  1453. # is the "migration barrier". A migration barrier of 1 means that a replica
  1454. # will migrate only if there is at least 1 other working replica for its master
  1455. # and so forth. It usually reflects the number of replicas you want for every
  1456. # master in your cluster.
  1457. #
  1458. # Default is 1 (replicas migrate only if their masters remain with at least
  1459. # one replica). To disable migration just set it to a very large value or
  1460. # set cluster-allow-replica-migration to 'no'.
  1461. # A value of 0 can be set but is useful only for debugging and dangerous
  1462. # in production.
  1463. #
  1464. # cluster-migration-barrier 1
  1466. # Turning off this option allows to use less automatic cluster configuration.
  1467. # It both disables migration to orphaned masters and migration from masters
  1468. # that became empty.
  1469. #
  1470. # Default is 'yes' (allow automatic migrations).
  1471. #
  1472. # cluster-allow-replica-migration yes
  1474. # By default Redis Cluster nodes stop accepting queries if they detect there
  1475. # is at least a hash slot uncovered (no available node is serving it).
  1476. # This way if the cluster is partially down (for example a range of hash slots
  1477. # are no longer covered) all the cluster becomes, eventually, unavailable.
  1478. # It automatically returns available as soon as all the slots are covered again.
  1479. #
  1480. # However sometimes you want the subset of the cluster which is working,
  1481. # to continue to accept queries for the part of the key space that is still
  1482. # covered. In order to do so, just set the cluster-require-full-coverage
  1483. # option to no.
  1484. #
  1485. # cluster-require-full-coverage yes
  1487. # This option, when set to yes, prevents replicas from trying to failover its
  1488. # master during master failures. However the replica can still perform a
  1489. # manual failover, if forced to do so.
  1490. #
  1491. # This is useful in different scenarios, especially in the case of multiple
  1492. # data center operations, where we want one side to never be promoted if not
  1493. # in the case of a total DC failure.
  1494. #
  1495. # cluster-replica-no-failover no
  1497. # This option, when set to yes, allows nodes to serve read traffic while the
  1498. # the cluster is in a down state, as long as it believes it owns the slots. 
  1499. #
  1500. # This is useful for two cases.  The first case is for when an application 
  1501. # doesn't require consistency of data during node failures or network partitions.
  1502. # One example of this is a cache, where as long as the node has the data it
  1503. # should be able to serve it. 
  1504. #
  1505. # The second use case is for configurations that don't meet the recommended  
  1506. # three shards but want to enable cluster mode and scale later. A 
  1507. # master outage in a 1 or 2 shard configuration causes a read/write outage to the
  1508. # entire cluster without this option set, with it set there is only a write outage.
  1509. # Without a quorum of masters, slot ownership will not change automatically. 
  1510. #
  1511. # cluster-allow-reads-when-down no
  1513. # In order to setup your cluster make sure to read the documentation
  1514. # available at https://redis.io web site.
  1516. ########################## CLUSTER DOCKER/NAT support  ########################
  1518. # In certain deployments, Redis Cluster nodes address discovery fails, because
  1519. # addresses are NAT-ted or because ports are forwarded (the typical case is
  1520. # Docker and other containers).
  1521. #
  1522. # In order to make Redis Cluster working in such environments, a static
  1523. # configuration where each node knows its public address is needed. The
  1524. # following four options are used for this scope, and are:
  1525. #
  1526. # * cluster-announce-ip
  1527. # * cluster-announce-port
  1528. # * cluster-announce-tls-port
  1529. # * cluster-announce-bus-port
  1530. #
  1531. # Each instructs the node about its address, client ports (for connections
  1532. # without and with TLS) and cluster message bus port. The information is then
  1533. # published in the header of the bus packets so that other nodes will be able to
  1534. # correctly map the address of the node publishing the information.
  1535. #
  1536. # If cluster-tls is set to yes and cluster-announce-tls-port is omitted or set
  1537. # to zero, then cluster-announce-port refers to the TLS port. Note also that
  1538. # cluster-announce-tls-port has no effect if cluster-tls is set to no.
  1539. #
  1540. # If the above options are not used, the normal Redis Cluster auto-detection
  1541. # will be used instead.
  1542. #
  1543. # Note that when remapped, the bus port may not be at the fixed offset of
  1544. # clients port + 10000, so you can specify any port and bus-port depending
  1545. # on how they get remapped. If the bus-port is not set, a fixed offset of
  1546. # 10000 will be used as usual.
  1547. #
  1548. # Example:
  1549. #
  1550. # cluster-announce-ip 10.1.1.5
  1551. # cluster-announce-tls-port 6379
  1552. # cluster-announce-port 0
  1553. # cluster-announce-bus-port 6380
  1555. ################################## SLOW LOG ###################################
  1557. # The Redis Slow Log is a system to log queries that exceeded a specified
  1558. # execution time. The execution time does not include the I/O operations
  1559. # like talking with the client, sending the reply and so forth,
  1560. # but just the time needed to actually execute the command (this is the only
  1561. # stage of command execution where the thread is blocked and can not serve
  1562. # other requests in the meantime).
  1563. #
  1564. # You can configure the slow log with two parameters: one tells Redis
  1565. # what is the execution time, in microseconds, to exceed in order for the
  1566. # command to get logged, and the other parameter is the length of the
  1567. # slow log. When a new command is logged the oldest one is removed from the
  1568. # queue of logged commands.
  1570. # The following time is expressed in microseconds, so 1000000 is equivalent
  1571. # to one second. Note that a negative number disables the slow log, while
  1572. # a value of zero forces the logging of every command.
  1573. slowlog-log-slower-than 10000
  1575. # There is no limit to this length. Just be aware that it will consume memory.
  1576. # You can reclaim memory used by the slow log with SLOWLOG RESET.
  1577. slowlog-max-len 128
  1579. ################################ LATENCY MONITOR ##############################
  1581. # The Redis latency monitoring subsystem samples different operations
  1582. # at runtime in order to collect data related to possible sources of
  1583. # latency of a Redis instance.
  1584. #
  1585. # Via the LATENCY command this information is available to the user that can
  1586. # print graphs and obtain reports.
  1587. #
  1588. # The system only logs operations that were performed in a time equal or
  1589. # greater than the amount of milliseconds specified via the
  1590. # latency-monitor-threshold configuration directive. When its value is set
  1591. # to zero, the latency monitor is turned off.
  1592. #
  1593. # By default latency monitoring is disabled since it is mostly not needed
  1594. # if you don't have latency issues, and collecting data has a performance
  1595. # impact, that while very small, can be measured under big load. Latency
  1596. # monitoring can easily be enabled at runtime using the command
  1597. # "CONFIG SET latency-monitor-threshold <milliseconds>" if needed.
  1598. latency-monitor-threshold 0
  1600. ############################# EVENT NOTIFICATION ##############################
  1602. # Redis can notify Pub/Sub clients about events happening in the key space.
  1603. # This feature is documented at https://redis.io/topics/notifications
  1604. #
  1605. # For instance if keyspace events notification is enabled, and a client
  1606. # performs a DEL operation on key "foo" stored in the Database 0, two
  1607. # messages will be published via Pub/Sub:
  1608. #
  1609. # PUBLISH __keyspace@0__:foo del
  1610. # PUBLISH __keyevent@0__:del foo
  1611. #
  1612. # It is possible to select the events that Redis will notify among a set
  1613. # of classes. Every class is identified by a single character:
  1614. #
  1615. #  K     Keyspace events, published with __keyspace@<db>__ prefix.
  1616. #  E     Keyevent events, published with __keyevent@<db>__ prefix.
  1617. #  g     Generic commands (non-type specific) like DEL, EXPIRE, RENAME, ...
  1618. #  $     String commands
  1619. #  l     List commands
  1620. #  s     Set commands
  1621. #  h     Hash commands
  1622. #  z     Sorted set commands
  1623. #  x     Expired events (events generated every time a key expires)
  1624. #  e     Evicted events (events generated when a key is evicted for maxmemory)
  1625. #  t     Stream commands
  1626. #  d     Module key type events
  1627. #  m     Key-miss events (Note: It is not included in the 'A' class)
  1628. #  A     Alias for g$lshzxetd, so that the "AKE" string means all the events
  1629. #        (Except key-miss events which are excluded from 'A' due to their
  1630. #         unique nature).
  1631. #
  1632. #  The "notify-keyspace-events" takes as argument a string that is composed
  1633. #  of zero or multiple characters. The empty string means that notifications
  1634. #  are disabled.
  1635. #
  1636. #  Example: to enable list and generic events, from the point of view of the
  1637. #           event name, use:
  1638. #
  1639. #  notify-keyspace-events Elg
  1640. #
  1641. #  Example 2: to get the stream of the expired keys subscribing to channel
  1642. #             name __keyevent@0__:expired use:
  1643. #
  1644. #  notify-keyspace-events Ex
  1645. #
  1646. #  By default all notifications are disabled because most users don't need
  1647. #  this feature and the feature has some overhead. Note that if you don't
  1648. #  specify at least one of K or E, no events will be delivered.
  1649. notify-keyspace-events ""
  1651. ############################### GOPHER SERVER #################################
  1653. # Redis contains an implementation of the Gopher protocol, as specified in
  1654. # the RFC 1436 (https://www.ietf.org/rfc/rfc1436.txt).
  1655. #
  1656. # The Gopher protocol was very popular in the late '90s. It is an alternative
  1657. # to the web, and the implementation both server and client side is so simple
  1658. # that the Redis server has just 100 lines of code in order to implement this
  1659. # support.
  1660. #
  1661. # What do you do with Gopher nowadays? Well Gopher never *really* died, and
  1662. # lately there is a movement in order for the Gopher more hierarchical content
  1663. # composed of just plain text documents to be resurrected. Some want a simpler
  1664. # internet, others believe that the mainstream internet became too much
  1665. # controlled, and it's cool to create an alternative space for people that
  1666. # want a bit of fresh air.
  1667. #
  1668. # Anyway for the 10nth birthday of the Redis, we gave it the Gopher protocol
  1669. # as a gift.
  1670. #
  1671. # --- HOW IT WORKS? ---
  1672. #
  1673. # The Redis Gopher support uses the inline protocol of Redis, and specifically
  1674. # two kind of inline requests that were anyway illegal: an empty request
  1675. # or any request that starts with "/" (there are no Redis commands starting
  1676. # with such a slash). Normal RESP2/RESP3 requests are completely out of the
  1677. # path of the Gopher protocol implementation and are served as usual as well.
  1678. #
  1679. # If you open a connection to Redis when Gopher is enabled and send it
  1680. # a string like "/foo", if there is a key named "/foo" it is served via the
  1681. # Gopher protocol.
  1682. #
  1683. # In order to create a real Gopher "hole" (the name of a Gopher site in Gopher
  1684. # talking), you likely need a script like the following:
  1685. #
  1686. #   https://github.com/antirez/gopher2redis
  1687. #
  1688. # --- SECURITY WARNING ---
  1689. #
  1690. # If you plan to put Redis on the internet in a publicly accessible address
  1691. # to server Gopher pages MAKE SURE TO SET A PASSWORD to the instance.
  1692. # Once a password is set:
  1693. #
  1694. #   1. The Gopher server (when enabled, not by default) will still serve
  1695. #      content via Gopher.
  1696. #   2. However other commands cannot be called before the client will
  1697. #      authenticate.
  1698. #
  1699. # So use the 'requirepass' option to protect your instance.
  1700. #
  1701. # Note that Gopher is not currently supported when 'io-threads-do-reads'
  1702. # is enabled.
  1703. #
  1704. # To enable Gopher support, uncomment the following line and set the option
  1705. # from no (the default) to yes.
  1706. #
  1707. # gopher-enabled no
  1709. ############################### ADVANCED CONFIG ###############################
  1711. # Hashes are encoded using a memory efficient data structure when they have a
  1712. # small number of entries, and the biggest entry does not exceed a given
  1713. # threshold. These thresholds can be configured using the following directives.
  1714. hash-max-ziplist-entries 512
  1715. hash-max-ziplist-value 64
  1717. # Lists are also encoded in a special way to save a lot of space.
  1718. # The number of entries allowed per internal list node can be specified
  1719. # as a fixed maximum size or a maximum number of elements.
  1720. # For a fixed maximum size, use -5 through -1, meaning:
  1721. # -5: max size: 64 Kb  <-- not recommended for normal workloads
  1722. # -4: max size: 32 Kb  <-- not recommended
  1723. # -3: max size: 16 Kb  <-- probably not recommended
  1724. # -2: max size: 8 Kb   <-- good
  1725. # -1: max size: 4 Kb   <-- good
  1726. # Positive numbers mean store up to _exactly_ that number of elements
  1727. # per list node.
  1728. # The highest performing option is usually -2 (8 Kb size) or -1 (4 Kb size),
  1729. # but if your use case is unique, adjust the settings as necessary.
  1730. list-max-ziplist-size -2
  1732. # Lists may also be compressed.
  1733. # Compress depth is the number of quicklist ziplist nodes from *each* side of
  1734. # the list to *exclude* from compression.  The head and tail of the list
  1735. # are always uncompressed for fast push/pop operations.  Settings are:
  1736. # 0: disable all list compression
  1737. # 1: depth 1 means "don't start compressing until after 1 node into the list,
  1738. #    going from either the head or tail"
  1739. #    So: [head]->node->node->...->node->[tail]
  1740. #    [head], [tail] will always be uncompressed; inner nodes will compress.
  1741. # 2: [head]->[next]->node->node->...->node->[prev]->[tail]
  1742. #    2 here means: don't compress head or head->next or tail->prev or tail,
  1743. #    but compress all nodes between them.
  1744. # 3: [head]->[next]->[next]->node->node->...->node->[prev]->[prev]->[tail]
  1745. # etc.
  1746. list-compress-depth 0
  1748. # Sets have a special encoding in just one case: when a set is composed
  1749. # of just strings that happen to be integers in radix 10 in the range
  1750. # of 64 bit signed integers.
  1751. # The following configuration setting sets the limit in the size of the
  1752. # set in order to use this special memory saving encoding.
  1753. set-max-intset-entries 512
  1755. # Similarly to hashes and lists, sorted sets are also specially encoded in
  1756. # order to save a lot of space. This encoding is only used when the length and
  1757. # elements of a sorted set are below the following limits:
  1758. zset-max-ziplist-entries 128
  1759. zset-max-ziplist-value 64
  1761. # HyperLogLog sparse representation bytes limit. The limit includes the
  1762. # 16 bytes header. When an HyperLogLog using the sparse representation crosses
  1763. # this limit, it is converted into the dense representation.
  1764. #
  1765. # A value greater than 16000 is totally useless, since at that point the
  1766. # dense representation is more memory efficient.
  1767. #
  1768. # The suggested value is ~ 3000 in order to have the benefits of
  1769. # the space efficient encoding without slowing down too much PFADD,
  1770. # which is O(N) with the sparse encoding. The value can be raised to
  1771. # ~ 10000 when CPU is not a concern, but space is, and the data set is
  1772. # composed of many HyperLogLogs with cardinality in the 0 - 15000 range.
  1773. hll-sparse-max-bytes 3000
  1775. # Streams macro node max size / items. The stream data structure is a radix
  1776. # tree of big nodes that encode multiple items inside. Using this configuration
  1777. # it is possible to configure how big a single node can be in bytes, and the
  1778. # maximum number of items it may contain before switching to a new node when
  1779. # appending new stream entries. If any of the following settings are set to
  1780. # zero, the limit is ignored, so for instance it is possible to set just a
  1781. # max entries limit by setting max-bytes to 0 and max-entries to the desired
  1782. # value.
  1783. stream-node-max-bytes 4096
  1784. stream-node-max-entries 100
  1786. # Active rehashing uses 1 millisecond every 100 milliseconds of CPU time in
  1787. # order to help rehashing the main Redis hash table (the one mapping top-level
  1788. # keys to values). The hash table implementation Redis uses (see dict.c)
  1789. # performs a lazy rehashing: the more operation you run into a hash table
  1790. # that is rehashing, the more rehashing "steps" are performed, so if the
  1791. # server is idle the rehashing is never complete and some more memory is used
  1792. # by the hash table.
  1793. #
  1794. # The default is to use this millisecond 10 times every second in order to
  1795. # actively rehash the main dictionaries, freeing memory when possible.
  1796. #
  1797. # If unsure:
  1798. # use "activerehashing no" if you have hard latency requirements and it is
  1799. # not a good thing in your environment that Redis can reply from time to time
  1800. # to queries with 2 milliseconds delay.
  1801. #
  1802. # use "activerehashing yes" if you don't have such hard requirements but
  1803. # want to free memory asap when possible.
  1804. activerehashing yes
  1806. # The client output buffer limits can be used to force disconnection of clients
  1807. # that are not reading data from the server fast enough for some reason (a
  1808. # common reason is that a Pub/Sub client can't consume messages as fast as the
  1809. # publisher can produce them).
  1810. #
  1811. # The limit can be set differently for the three different classes of clients:
  1812. #
  1813. # normal -> normal clients including MONITOR clients
  1814. # replica  -> replica clients
  1815. # pubsub -> clients subscribed to at least one pubsub channel or pattern
  1816. #
  1817. # The syntax of every client-output-buffer-limit directive is the following:
  1818. #
  1819. # client-output-buffer-limit <class> <hard limit> <soft limit> <soft seconds>
  1820. #
  1821. # A client is immediately disconnected once the hard limit is reached, or if
  1822. # the soft limit is reached and remains reached for the specified number of
  1823. # seconds (continuously).
  1824. # So for instance if the hard limit is 32 megabytes and the soft limit is
  1825. # 16 megabytes / 10 seconds, the client will get disconnected immediately
  1826. # if the size of the output buffers reach 32 megabytes, but will also get
  1827. # disconnected if the client reaches 16 megabytes and continuously overcomes
  1828. # the limit for 10 seconds.
  1829. #
  1830. # By default normal clients are not limited because they don't receive data
  1831. # without asking (in a push way), but just after a request, so only
  1832. # asynchronous clients may create a scenario where data is requested faster
  1833. # than it can read.
  1834. #
  1835. # Instead there is a default limit for pubsub and replica clients, since
  1836. # subscribers and replicas receive data in a push fashion.
  1837. #
  1838. # Both the hard or the soft limit can be disabled by setting them to zero.
  1839. client-output-buffer-limit normal 0 0 0
  1840. client-output-buffer-limit replica 256mb 64mb 60
  1841. client-output-buffer-limit pubsub 32mb 8mb 60
  1843. # Client query buffers accumulate new commands. They are limited to a fixed
  1844. # amount by default in order to avoid that a protocol desynchronization (for
  1845. # instance due to a bug in the client) will lead to unbound memory usage in
  1846. # the query buffer. However you can configure it here if you have very special
  1847. # needs, such us huge multi/exec requests or alike.
  1848. #
  1849. # client-query-buffer-limit 1gb
  1851. # In the Redis protocol, bulk requests, that are, elements representing single
  1852. # strings, are normally limited to 512 mb. However you can change this limit
  1853. # here, but must be 1mb or greater
  1854. #
  1855. # proto-max-bulk-len 512mb
  1857. # Redis calls an internal function to perform many background tasks, like
  1858. # closing connections of clients in timeout, purging expired keys that are
  1859. # never requested, and so forth.
  1860. #
  1861. # Not all tasks are performed with the same frequency, but Redis checks for
  1862. # tasks to perform according to the specified "hz" value.
  1863. #
  1864. # By default "hz" is set to 10. Raising the value will use more CPU when
  1865. # Redis is idle, but at the same time will make Redis more responsive when
  1866. # there are many keys expiring at the same time, and timeouts may be
  1867. # handled with more precision.
  1868. #
  1869. # The range is between 1 and 500, however a value over 100 is usually not
  1870. # a good idea. Most users should use the default of 10 and raise this up to
  1871. # 100 only in environments where very low latency is required.
  1872. hz 10
  1874. # Normally it is useful to have an HZ value which is proportional to the
  1875. # number of clients connected. This is useful in order, for instance, to
  1876. # avoid too many clients are processed for each background task invocation
  1877. # in order to avoid latency spikes.
  1878. #
  1879. # Since the default HZ value by default is conservatively set to 10, Redis
  1880. # offers, and enables by default, the ability to use an adaptive HZ value
  1881. # which will temporarily raise when there are many connected clients.
  1882. #
  1883. # When dynamic HZ is enabled, the actual configured HZ will be used
  1884. # as a baseline, but multiples of the configured HZ value will be actually
  1885. # used as needed once more clients are connected. In this way an idle
  1886. # instance will use very little CPU time while a busy instance will be
  1887. # more responsive.
  1888. dynamic-hz yes
  1890. # When a child rewrites the AOF file, if the following option is enabled
  1891. # the file will be fsync-ed every 32 MB of data generated. This is useful
  1892. # in order to commit the file to the disk more incrementally and avoid
  1893. # big latency spikes.
  1894. aof-rewrite-incremental-fsync yes
  1896. # When redis saves RDB file, if the following option is enabled
  1897. # the file will be fsync-ed every 32 MB of data generated. This is useful
  1898. # in order to commit the file to the disk more incrementally and avoid
  1899. # big latency spikes.
  1900. rdb-save-incremental-fsync yes
  1902. # Redis LFU eviction (see maxmemory setting) can be tuned. However it is a good
  1903. # idea to start with the default settings and only change them after investigating
  1904. # how to improve the performances and how the keys LFU change over time, which
  1905. # is possible to inspect via the OBJECT FREQ command.
  1906. #
  1907. # There are two tunable parameters in the Redis LFU implementation: the
  1908. # counter logarithm factor and the counter decay time. It is important to
  1909. # understand what the two parameters mean before changing them.
  1910. #
  1911. # The LFU counter is just 8 bits per key, it's maximum value is 255, so Redis
  1912. # uses a probabilistic increment with logarithmic behavior. Given the value
  1913. # of the old counter, when a key is accessed, the counter is incremented in
  1914. # this way:
  1915. #
  1916. # 1. A random number R between 0 and 1 is extracted.
  1917. # 2. A probability P is calculated as 1/(old_value*lfu_log_factor+1).
  1918. # 3. The counter is incremented only if R < P.
  1919. #
  1920. # The default lfu-log-factor is 10. This is a table of how the frequency
  1921. # counter changes with a different number of accesses with different
  1922. # logarithmic factors:
  1923. #
  1924. # +--------+------------+------------+------------+------------+------------+
  1925. # | factor | 100 hits   | 1000 hits  | 100K hits  | 1M hits    | 10M hits   |
  1926. # +--------+------------+------------+------------+------------+------------+
  1927. # | 0      | 104        | 255        | 255        | 255        | 255        |
  1928. # +--------+------------+------------+------------+------------+------------+
  1929. # | 1      | 18         | 49         | 255        | 255        | 255        |
  1930. # +--------+------------+------------+------------+------------+------------+
  1931. # | 10     | 10         | 18         | 142        | 255        | 255        |
  1932. # +--------+------------+------------+------------+------------+------------+
  1933. # | 100    | 8          | 11         | 49         | 143        | 255        |
  1934. # +--------+------------+------------+------------+------------+------------+
  1935. #
  1936. # NOTE: The above table was obtained by running the following commands:
  1937. #
  1938. #   redis-benchmark -n 1000000 incr foo
  1939. #   redis-cli object freq foo
  1940. #
  1941. # NOTE 2: The counter initial value is 5 in order to give new objects a chance
  1942. # to accumulate hits.
  1943. #
  1944. # The counter decay time is the time, in minutes, that must elapse in order
  1945. # for the key counter to be divided by two (or decremented if it has a value
  1946. # less <= 10).
  1947. #
  1948. # The default value for the lfu-decay-time is 1. A special value of 0 means to
  1949. # decay the counter every time it happens to be scanned.
  1950. #
  1951. # lfu-log-factor 10
  1952. # lfu-decay-time 1
  1954. ########################### ACTIVE DEFRAGMENTATION #######################
  1955. #
  1956. # What is active defragmentation?
  1957. # -------------------------------
  1958. #
  1959. # Active (online) defragmentation allows a Redis server to compact the
  1960. # spaces left between small allocations and deallocations of data in memory,
  1961. # thus allowing to reclaim back memory.
  1962. #
  1963. # Fragmentation is a natural process that happens with every allocator (but
  1964. # less so with Jemalloc, fortunately) and certain workloads. Normally a server
  1965. # restart is needed in order to lower the fragmentation, or at least to flush
  1966. # away all the data and create it again. However thanks to this feature
  1967. # implemented by Oran Agra for Redis 4.0 this process can happen at runtime
  1968. # in a "hot" way, while the server is running.
  1969. #
  1970. # Basically when the fragmentation is over a certain level (see the
  1971. # configuration options below) Redis will start to create new copies of the
  1972. # values in contiguous memory regions by exploiting certain specific Jemalloc
  1973. # features (in order to understand if an allocation is causing fragmentation
  1974. # and to allocate it in a better place), and at the same time, will release the
  1975. # old copies of the data. This process, repeated incrementally for all the keys
  1976. # will cause the fragmentation to drop back to normal values.
  1977. #
  1978. # Important things to understand:
  1979. #
  1980. # 1. This feature is disabled by default, and only works if you compiled Redis
  1981. #    to use the copy of Jemalloc we ship with the source code of Redis.
  1982. #    This is the default with Linux builds.
  1983. #
  1984. # 2. You never need to enable this feature if you don't have fragmentation
  1985. #    issues.
  1986. #
  1987. # 3. Once you experience fragmentation, you can enable this feature when
  1988. #    needed with the command "CONFIG SET activedefrag yes".
  1989. #
  1990. # The configuration parameters are able to fine tune the behavior of the
  1991. # defragmentation process. If you are not sure about what they mean it is
  1992. # a good idea to leave the defaults untouched.
  1994. # Enabled active defragmentation
  1995. # activedefrag no
  1997. # Minimum amount of fragmentation waste to start active defrag
  1998. # active-defrag-ignore-bytes 100mb
  2000. # Minimum percentage of fragmentation to start active defrag
  2001. # active-defrag-threshold-lower 10
  2003. # Maximum percentage of fragmentation at which we use maximum effort
  2004. # active-defrag-threshold-upper 100
  2006. # Minimal effort for defrag in CPU percentage, to be used when the lower
  2007. # threshold is reached
  2008. # active-defrag-cycle-min 1
  2010. # Maximal effort for defrag in CPU percentage, to be used when the upper
  2011. # threshold is reached
  2012. # active-defrag-cycle-max 25
  2014. # Maximum number of set/hash/zset/list fields that will be processed from
  2015. # the main dictionary scan
  2016. # active-defrag-max-scan-fields 1000
  2018. # Jemalloc background thread for purging will be enabled by default
  2019. jemalloc-bg-thread yes
  2021. # It is possible to pin different threads and processes of Redis to specific
  2022. # CPUs in your system, in order to maximize the performances of the server.
  2023. # This is useful both in order to pin different Redis threads in different
  2024. # CPUs, but also in order to make sure that multiple Redis instances running
  2025. # in the same host will be pinned to different CPUs.
  2026. #
  2027. # Normally you can do this using the "taskset" command, however it is also
  2028. # possible to this via Redis configuration directly, both in Linux and FreeBSD.
  2029. #
  2030. # You can pin the server/IO threads, bio threads, aof rewrite child process, and
  2031. # the bgsave child process. The syntax to specify the cpu list is the same as
  2032. # the taskset command:
  2033. #
  2034. # Set redis server/io threads to cpu affinity 0,2,4,6:
  2035. # server_cpulist 0-7:2
  2036. #
  2037. # Set bio threads to cpu affinity 1,3:
  2038. # bio_cpulist 1,3
  2039. #
  2040. # Set aof rewrite child process to cpu affinity 8,9,10,11:
  2041. # aof_rewrite_cpulist 8-11
  2042. #
  2043. # Set bgsave child process to cpu affinity 1,10,11
  2044. # bgsave_cpulist 1,10-11
  2046. # In some cases redis will emit warnings and even refuse to start if it detects
  2047. # that the system is in bad state, it is possible to suppress these warnings
  2048. # by setting the following config which takes a space delimited list of warnings
  2049. # to suppress
  2050. #
  2051. # ignore-warnings ARM64-COW-BUG