1This README is just a fast *quick start* document. You can find more detailed documentation at [redis.io](https://redis.io). 2 3What is Redis? 4-------------- 5 6Redis is often referred as a *data structures* server. What this means is that Redis provides access to mutable data structures via a set of commands, which are sent using a *server-client* model with TCP sockets and a simple protocol. So different processes can query and modify the same data structures in a shared way. 7 8Data structures implemented into Redis have a few special properties: 9 10* Redis cares to store them on disk, even if they are always served and modified into the server memory. This means that Redis is fast, but that is also non-volatile. 11* Implementation of data structures stress on memory efficiency, so data structures inside Redis will likely use less memory compared to the same data structure modeled using an high level programming language. 12* Redis offers a number of features that are natural to find in a database, like replication, tunable levels of durability, cluster, high availability. 13 14Another good example is to think of Redis as a more complex version of memcached, where the operations are not just SETs and GETs, but operations to work with complex data types like Lists, Sets, ordered data structures, and so forth. 15 16If you want to know more, this is a list of selected starting points: 17 18* Introduction to Redis data types. http://redis.io/topics/data-types-intro 19* Try Redis directly inside your browser. http://try.redis.io 20* The full list of Redis commands. http://redis.io/commands 21* There is much more inside the Redis official documentation. http://redis.io/documentation 22 23Building Redis 24-------------- 25 26Redis can be compiled and used on Linux, OSX, OpenBSD, NetBSD, FreeBSD. 27We support big endian and little endian architectures, and both 32 bit 28and 64 bit systems. 29 30It may compile on Solaris derived systems (for instance SmartOS) but our 31support for this platform is *best effort* and Redis is not guaranteed to 32work as well as in Linux, OSX, and \*BSD there. 33 34It is as simple as: 35 36 % make 37 38You can run a 32 bit Redis binary using: 39 40 % make 32bit 41 42After building Redis, it is a good idea to test it using: 43 44 % make test 45 46Fixing build problems with dependencies or cached build options 47--------- 48 49Redis has some dependencies which are included into the `deps` directory. 50`make` does not automatically rebuild dependencies even if something in 51the source code of dependencies changes. 52 53When you update the source code with `git pull` or when code inside the 54dependencies tree is modified in any other way, make sure to use the following 55command in order to really clean everything and rebuild from scratch: 56 57 make distclean 58 59This will clean: jemalloc, lua, hiredis, linenoise. 60 61Also if you force certain build options like 32bit target, no C compiler 62optimizations (for debugging purposes), and other similar build time options, 63those options are cached indefinitely until you issue a `make distclean` 64command. 65 66Fixing problems building 32 bit binaries 67--------- 68 69If after building Redis with a 32 bit target you need to rebuild it 70with a 64 bit target, or the other way around, you need to perform a 71`make distclean` in the root directory of the Redis distribution. 72 73In case of build errors when trying to build a 32 bit binary of Redis, try 74the following steps: 75 76* Install the packages libc6-dev-i386 (also try g++-multilib). 77* Try using the following command line instead of `make 32bit`: 78 `make CFLAGS="-m32 -march=native" LDFLAGS="-m32"` 79 80Allocator 81--------- 82 83Selecting a non-default memory allocator when building Redis is done by setting 84the `MALLOC` environment variable. Redis is compiled and linked against libc 85malloc by default, with the exception of jemalloc being the default on Linux 86systems. This default was picked because jemalloc has proven to have fewer 87fragmentation problems than libc malloc. 88 89To force compiling against libc malloc, use: 90 91 % make MALLOC=libc 92 93To compile against jemalloc on Mac OS X systems, use: 94 95 % make MALLOC=jemalloc 96 97Verbose build 98------------- 99 100Redis will build with a user friendly colorized output by default. 101If you want to see a more verbose output use the following: 102 103 % make V=1 104 105Running Redis 106------------- 107 108To run Redis with the default configuration just type: 109 110 % cd src 111 % ./redis-server 112 113If you want to provide your redis.conf, you have to run it using an additional 114parameter (the path of the configuration file): 115 116 % cd src 117 % ./redis-server /path/to/redis.conf 118 119It is possible to alter the Redis configuration by passing parameters directly 120as options using the command line. Examples: 121 122 % ./redis-server --port 9999 --replicaof 127.0.0.1 6379 123 % ./redis-server /etc/redis/6379.conf --loglevel debug 124 125All the options in redis.conf are also supported as options using the command 126line, with exactly the same name. 127 128Playing with Redis 129------------------ 130 131You can use redis-cli to play with Redis. Start a redis-server instance, 132then in another terminal try the following: 133 134 % cd src 135 % ./redis-cli 136 redis> ping 137 PONG 138 redis> set foo bar 139 OK 140 redis> get foo 141 "bar" 142 redis> incr mycounter 143 (integer) 1 144 redis> incr mycounter 145 (integer) 2 146 redis> 147 148You can find the list of all the available commands at http://redis.io/commands. 149 150Installing Redis 151----------------- 152 153In order to install Redis binaries into /usr/local/bin just use: 154 155 % make install 156 157You can use `make PREFIX=/some/other/directory install` if you wish to use a 158different destination. 159 160Make install will just install binaries in your system, but will not configure 161init scripts and configuration files in the appropriate place. This is not 162needed if you want just to play a bit with Redis, but if you are installing 163it the proper way for a production system, we have a script doing this 164for Ubuntu and Debian systems: 165 166 % cd utils 167 % ./install_server.sh 168 169The script will ask you a few questions and will setup everything you need 170to run Redis properly as a background daemon that will start again on 171system reboots. 172 173You'll be able to stop and start Redis using the script named 174`/etc/init.d/redis_<portnumber>`, for instance `/etc/init.d/redis_6379`. 175 176Code contributions 177----------------- 178 179Note: by contributing code to the Redis project in any form, including sending 180a pull request via Github, a code fragment or patch via private email or 181public discussion groups, you agree to release your code under the terms 182of the BSD license that you can find in the [COPYING][1] file included in the Redis 183source distribution. 184 185Please see the [CONTRIBUTING][2] file in this source distribution for more 186information. 187 188[1]: https://github.com/antirez/redis/blob/unstable/COPYING 189[2]: https://github.com/antirez/redis/blob/unstable/CONTRIBUTING 190 191Redis internals 192=== 193 194If you are reading this README you are likely in front of a Github page 195or you just untarred the Redis distribution tar ball. In both the cases 196you are basically one step away from the source code, so here we explain 197the Redis source code layout, what is in each file as a general idea, the 198most important functions and structures inside the Redis server and so forth. 199We keep all the discussion at a high level without digging into the details 200since this document would be huge otherwise and our code base changes 201continuously, but a general idea should be a good starting point to 202understand more. Moreover most of the code is heavily commented and easy 203to follow. 204 205Source code layout 206--- 207 208The Redis root directory just contains this README, the Makefile which 209calls the real Makefile inside the `src` directory and an example 210configuration for Redis and Sentinel. You can find a few shell 211scripts that are used in order to execute the Redis, Redis Cluster and 212Redis Sentinel unit tests, which are implemented inside the `tests` 213directory. 214 215Inside the root are the following important directories: 216 217* `src`: contains the Redis implementation, written in C. 218* `tests`: contains the unit tests, implemented in Tcl. 219* `deps`: contains libraries Redis uses. Everything needed to compile Redis is inside this directory; your system just needs to provide `libc`, a POSIX compatible interface and a C compiler. Notably `deps` contains a copy of `jemalloc`, which is the default allocator of Redis under Linux. Note that under `deps` there are also things which started with the Redis project, but for which the main repository is not `antirez/redis`. An exception to this rule is `deps/geohash-int` which is the low level geocoding library used by Redis: it originated from a different project, but at this point it diverged so much that it is developed as a separated entity directly inside the Redis repository. 220 221There are a few more directories but they are not very important for our goals 222here. We'll focus mostly on `src`, where the Redis implementation is contained, 223exploring what there is inside each file. The order in which files are 224exposed is the logical one to follow in order to disclose different layers 225of complexity incrementally. 226 227Note: lately Redis was refactored quite a bit. Function names and file 228names have been changed, so you may find that this documentation reflects the 229`unstable` branch more closely. For instance in Redis 3.0 the `server.c` 230and `server.h` files were named `redis.c` and `redis.h`. However the overall 231structure is the same. Keep in mind that all the new developments and pull 232requests should be performed against the `unstable` branch. 233 234server.h 235--- 236 237The simplest way to understand how a program works is to understand the 238data structures it uses. So we'll start from the main header file of 239Redis, which is `server.h`. 240 241All the server configuration and in general all the shared state is 242defined in a global structure called `server`, of type `struct redisServer`. 243A few important fields in this structure are: 244 245* `server.db` is an array of Redis databases, where data is stored. 246* `server.commands` is the command table. 247* `server.clients` is a linked list of clients connected to the server. 248* `server.master` is a special client, the master, if the instance is a replica. 249 250There are tons of other fields. Most fields are commented directly inside 251the structure definition. 252 253Another important Redis data structure is the one defining a client. 254In the past it was called `redisClient`, now just `client`. The structure 255has many fields, here we'll just show the main ones: 256 257 struct client { 258 int fd; 259 sds querybuf; 260 int argc; 261 robj **argv; 262 redisDb *db; 263 int flags; 264 list *reply; 265 char buf[PROTO_REPLY_CHUNK_BYTES]; 266 ... many other fields ... 267 } 268 269The client structure defines a *connected client*: 270 271* The `fd` field is the client socket file descriptor. 272* `argc` and `argv` are populated with the command the client is executing, so that functions implementing a given Redis command can read the arguments. 273* `querybuf` accumulates the requests from the client, which are parsed by the Redis server according to the Redis protocol and executed by calling the implementations of the commands the client is executing. 274* `reply` and `buf` are dynamic and static buffers that accumulate the replies the server sends to the client. These buffers are incrementally written to the socket as soon as the file descriptor is writable. 275 276As you can see in the client structure above, arguments in a command 277are described as `robj` structures. The following is the full `robj` 278structure, which defines a *Redis object*: 279 280 typedef struct redisObject { 281 unsigned type:4; 282 unsigned encoding:4; 283 unsigned lru:LRU_BITS; /* lru time (relative to server.lruclock) */ 284 int refcount; 285 void *ptr; 286 } robj; 287 288Basically this structure can represent all the basic Redis data types like 289strings, lists, sets, sorted sets and so forth. The interesting thing is that 290it has a `type` field, so that it is possible to know what type a given 291object has, and a `refcount`, so that the same object can be referenced 292in multiple places without allocating it multiple times. Finally the `ptr` 293field points to the actual representation of the object, which might vary 294even for the same type, depending on the `encoding` used. 295 296Redis objects are used extensively in the Redis internals, however in order 297to avoid the overhead of indirect accesses, recently in many places 298we just use plain dynamic strings not wrapped inside a Redis object. 299 300server.c 301--- 302 303This is the entry point of the Redis server, where the `main()` function 304is defined. The following are the most important steps in order to startup 305the Redis server. 306 307* `initServerConfig()` setups the default values of the `server` structure. 308* `initServer()` allocates the data structures needed to operate, setup the listening socket, and so forth. 309* `aeMain()` starts the event loop which listens for new connections. 310 311There are two special functions called periodically by the event loop: 312 3131. `serverCron()` is called periodically (according to `server.hz` frequency), and performs tasks that must be performed from time to time, like checking for timedout clients. 3142. `beforeSleep()` is called every time the event loop fired, Redis served a few requests, and is returning back into the event loop. 315 316Inside server.c you can find code that handles other vital things of the Redis server: 317 318* `call()` is used in order to call a given command in the context of a given client. 319* `activeExpireCycle()` handles eviciton of keys with a time to live set via the `EXPIRE` command. 320* `freeMemoryIfNeeded()` is called when a new write command should be performed but Redis is out of memory according to the `maxmemory` directive. 321* The global variable `redisCommandTable` defines all the Redis commands, specifying the name of the command, the function implementing the command, the number of arguments required, and other properties of each command. 322 323networking.c 324--- 325 326This file defines all the I/O functions with clients, masters and replicas 327(which in Redis are just special clients): 328 329* `createClient()` allocates and initializes a new client. 330* the `addReply*()` family of functions are used by commands implementations in order to append data to the client structure, that will be transmitted to the client as a reply for a given command executed. 331* `writeToClient()` transmits the data pending in the output buffers to the client and is called by the *writable event handler* `sendReplyToClient()`. 332* `readQueryFromClient()` is the *readable event handler* and accumulates data from read from the client into the query buffer. 333* `processInputBuffer()` is the entry point in order to parse the client query buffer according to the Redis protocol. Once commands are ready to be processed, it calls `processCommand()` which is defined inside `server.c` in order to actually execute the command. 334* `freeClient()` deallocates, disconnects and removes a client. 335 336aof.c and rdb.c 337--- 338 339As you can guess from the names these files implement the RDB and AOF 340persistence for Redis. Redis uses a persistence model based on the `fork()` 341system call in order to create a thread with the same (shared) memory 342content of the main Redis thread. This secondary thread dumps the content 343of the memory on disk. This is used by `rdb.c` to create the snapshots 344on disk and by `aof.c` in order to perform the AOF rewrite when the 345append only file gets too big. 346 347The implementation inside `aof.c` has additional functions in order to 348implement an API that allows commands to append new commands into the AOF 349file as clients execute them. 350 351The `call()` function defined inside `server.c` is responsible to call 352the functions that in turn will write the commands into the AOF. 353 354db.c 355--- 356 357Certain Redis commands operate on specific data types, others are general. 358Examples of generic commands are `DEL` and `EXPIRE`. They operate on keys 359and not on their values specifically. All those generic commands are 360defined inside `db.c`. 361 362Moreover `db.c` implements an API in order to perform certain operations 363on the Redis dataset without directly accessing the internal data structures. 364 365The most important functions inside `db.c` which are used in many commands 366implementations are the following: 367 368* `lookupKeyRead()` and `lookupKeyWrite()` are used in order to get a pointer to the value associated to a given key, or `NULL` if the key does not exist. 369* `dbAdd()` and its higher level counterpart `setKey()` create a new key in a Redis database. 370* `dbDelete()` removes a key and its associated value. 371* `emptyDb()` removes an entire single database or all the databases defined. 372 373The rest of the file implements the generic commands exposed to the client. 374 375object.c 376--- 377 378The `robj` structure defining Redis objects was already described. Inside 379`object.c` there are all the functions that operate with Redis objects at 380a basic level, like functions to allocate new objects, handle the reference 381counting and so forth. Notable functions inside this file: 382 383* `incrRefcount()` and `decrRefCount()` are used in order to increment or decrement an object reference count. When it drops to 0 the object is finally freed. 384* `createObject()` allocates a new object. There are also specialized functions to allocate string objects having a specific content, like `createStringObjectFromLongLong()` and similar functions. 385 386This file also implements the `OBJECT` command. 387 388replication.c 389--- 390 391This is one of the most complex files inside Redis, it is recommended to 392approach it only after getting a bit familiar with the rest of the code base. 393In this file there is the implementation of both the master and replica role 394of Redis. 395 396One of the most important functions inside this file is `replicationFeedSlaves()` that writes commands to the clients representing replica instances connected 397to our master, so that the replicas can get the writes performed by the clients: 398this way their data set will remain synchronized with the one in the master. 399 400This file also implements both the `SYNC` and `PSYNC` commands that are 401used in order to perform the first synchronization between masters and 402replicas, or to continue the replication after a disconnection. 403 404Other C files 405--- 406 407* `t_hash.c`, `t_list.c`, `t_set.c`, `t_string.c` and `t_zset.c` contains the implementation of the Redis data types. They implement both an API to access a given data type, and the client commands implementations for these data types. 408* `ae.c` implements the Redis event loop, it's a self contained library which is simple to read and understand. 409* `sds.c` is the Redis string library, check http://github.com/antirez/sds for more information. 410* `anet.c` is a library to use POSIX networking in a simpler way compared to the raw interface exposed by the kernel. 411* `dict.c` is an implementation of a non-blocking hash table which rehashes incrementally. 412* `scripting.c` implements Lua scripting. It is completely self contained from the rest of the Redis implementation and is simple enough to understand if you are familar with the Lua API. 413* `cluster.c` implements the Redis Cluster. Probably a good read only after being very familiar with the rest of the Redis code base. If you want to read `cluster.c` make sure to read the [Redis Cluster specification][3]. 414 415[3]: http://redis.io/topics/cluster-spec 416 417Anatomy of a Redis command 418--- 419 420All the Redis commands are defined in the following way: 421 422 void foobarCommand(client *c) { 423 printf("%s",c->argv[1]->ptr); /* Do something with the argument. */ 424 addReply(c,shared.ok); /* Reply something to the client. */ 425 } 426 427The command is then referenced inside `server.c` in the command table: 428 429 {"foobar",foobarCommand,2,"rtF",0,NULL,0,0,0,0,0}, 430 431In the above example `2` is the number of arguments the command takes, 432while `"rtF"` are the command flags, as documented in the command table 433top comment inside `server.c`. 434 435After the command operates in some way, it returns a reply to the client, 436usually using `addReply()` or a similar function defined inside `networking.c`. 437 438There are tons of commands implementations inside the Redis source code 439that can serve as examples of actual commands implementations. To write 440a few toy commands can be a good exercise to familiarize with the code base. 441 442There are also many other files not described here, but it is useless to 443cover everything. We want to just help you with the first steps. 444Eventually you'll find your way inside the Redis code base :-) 445 446Enjoy! 447