<% set_title("Redis API for", product_name) %>

<!--
Licensed to the Apache Software Foundation (ASF) under one or more
contributor license agreements.  See the NOTICE file distributed with
this work for additional information regarding copyright ownership.
The ASF licenses this file to You under the Apache License, Version 2.0
(the "License"); you may not use this file except in compliance with
the License.  You may obtain a copy of the License at

     http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
-->

The Redis API for <%=vars.product_name%> allows <%=vars.product_name%> to function as a drop-in replacement for a
highly-available Redis data store, letting Redis applications take advantage of
<%=vars.product_name%>’s scaling capabilities without changing their client code. Redis clients connect to a <%=vars.product_name%>
server in the same way they connect to a Redis server, using a hostname and a port number, with
optional password authentication.

<img src="../images/redis_api_for_geode.png" class="image" />

## <a id="using-the-api" class="no-quick-link"></a>Using the Redis API for <%=vars.product_name%>

The <%=vars.product_name%> cluster must have at least one server that is set up to handle the incoming Redis
commands.

Use gfsh to start at least one server with a command of the form:

```pre
start server \
  --name=<serverName> \
  --locators=<locatorPort> \
  --redis-port=<redisPort> \
  --redis-bind-address=<redisBindAddress> \
  --redis-password=<redisPassword>
```

Replace `<serverName>` with the name of your server.

Replace `<locatorPort>` with your locator port.

Replace `<redisPort>` with the port that the <%=vars.product_name%> server listens on for Redis commands. The typical port used with a Redis cluster is 6379.

Replace `<redisBindAddress>` with the address of the server host.

Replace `<redisPassword>` with the password clients use to authenticate.

To confirm the server is listening, run:

``` pre
redis-cli -h <redisBindAddress> -p <redisPort> -a <redisPassword> ping
```

Replace `<redisBindAddress>`,`<redisPort>`, and `<redisPassword>` with the same values as the server.

If the server is functioning properly, you should see a response of `PONG`.

## <a id="included-commands" class="no-quick-link"></a>Included Redis Commands

The Redis API for <%=vars.product_name%> currently supports the following commands.  See [Redis commands](https://redis.io/commands/) for a complete list of and more information on Redis commands.

**Note**: These commands are supported for Redis 5.

-   **Connection**: AUTH, PING, QUIT
-   **Hashes**: HGETALL, HMSET, HSET
-   **Keys**: DEL, EXISTS, EXPIRE, EXPIREAT, KEYS, PERSIST, PEXPIRE, PEXPIREAT, PTTL, RENAME, TTL, TYPE
-   **Publish/Subscribe**: PUBLISH, PSUBSCRIBE, PUNSUBSCRIBE, SUBSCRIBE, UNSUBSCRIBE
-   **Sets**: SADD, SMEMBERS, SREM
-   **Strings**: APPEND, GET, SET

The following Redis API for <%=vars.product_name%> commands are **unsupported**. Unsupported
commands are available to use, but have not been fully tested. There is no guarantee they will work
exactly as expected.

-   **Connection**: ECHO, SELECT
-   **Hashes**: HDEL, HEXISTS, HGET, HINCRBY, HINCRBYFLOAT, HKEYS, HLEN, HMGET, HSCAN, HSETNX, HVALS
-   **Keys**: SCAN, UNLINK
-   **Server**: DBSIZE, FLUSHALL (no async option), FLUSHDB (no async option), SHUTDOWN, SLOWLOG,
  TIME
-   **Sets**: SCARD, SDIFF, SDIFFSTORE, SINTER, SINTERSTORE, SISMEMBER, SMOVE, SPOP, SRANDMEMBER,
  SSCAN, SUNION, SUNIONSTORE
-   **Strings**: BITCOUNT, BITOP, BITPOS, DECR, DECRBY, GETBIT, GETRANGE, GETSET, INCR, INCRBY, INCRBYFLOAT, MGET,
  MSET, MSETNX, PSETEX, SETBIT, SETEX, SETNX, SETRANGE, STRLEN


If you already have some Geode servers running with Redis enabled, you can execute the following
command with gfsh to enable unsupported commands:

```pre
redis --enable-unsupported-commands
```

You can also enable unsupported commands when you start the Geode server by setting the Java property `enable-redis-unsupported-commands=true`:

```pre
start server \
  --J=-Denable-redis-unsupported-commands=true \
  --name=<serverName> \
  --locators=<locatorPort> \
  --redis-port=<redisPort> \
  --redis-bind-address=<redisBindAddress> \
  --redis-password=<redisPassword>
```

Commands not listed above are **not implemented**.

## <a id="advantages-over-redis" class="no-quick-link"></a>Advantages of <%=vars.product_name%> over Redis

<%=vars.product_name%>’s primary advantage is its **scalability**. While the Redis server is single threaded, <%=vars.product_name%> supports high concurrency. Many Redis clients can execute commands on the <%=vars.product_name%> cluster simultaneously.

<%=vars.product_name%>'s architecture and management features help detect and resolve **network partitioning** problems without explicit management on the part of the Redis client.

## <a id="expiration-accuracy" class="no-quick-link"></a>Expiration Accuracy

Keys are expired in two ways, actively and passively:

-   With active expiration, expiration is evaluated whenever a key is accessed. If the key is due to expire, it is deleted. Active expiration is accurate to the millisecond.
-   With passive expiration, keys are evaluated every second. If they are due to expire, they are deleted.  Passive expiration is accurate to the second.

## <a id="high-availability-model" class="no-quick-link"></a>High Availability Model

Data is stored in a single partitioned region that has one redundant copy.
In practice this means that the cluster can tolerate the loss of a single server without the loss of
data.

## <a id="loss-of-connections" class="no-quick-link"></a>Loss of Connections

There are a number of events that might occur within the <%=vars.product_name%> cluster that can result
in the cluster closing the connection to the Redis client. Losing the connection to the cluster does not
imply that the server is no longer available.

When the connection is lost, the client should attempt to reconnect to the same server before
attempting to connect to another server.
The Redis client is responsible for knowing the addresses of all servers.

In the case of a connection failure, an invoked command may or may not complete.
The Redis client is responsible for deciding if the command should be retried.