Caching and Redis

Mohammad Iqbal
Content
• What is Caching?
• Why caching?
• Architecture
• Type of Caching
• In Memory Cache
• Distributed Cache
• Redis
• Redis stand alone
• Redis Sentinel
• Redis Cluster
 What is Caching?

● A cache is a hardware or software component that stores Interview Questions

data so that future requests for that data can be served
faster. 1. What’s 2*2
happy me: 4
● The data will be either a pre-computed data or a result of
previous queries.
2. What’s 1578*1298*12312.
● The main advantage, and also the goal, of caching is Me: you are a good
speeding up any process API call, page loading etc.. question

● Increase the overall cost of the application but the trade of

between the cost and feature is always beneficial.
Why Caching?

● Caching data is important because it helps speed

up the response time or wait time for user.
● As a large number of queries can be served by
cache it reduces overall load on servers which
increase efficiency.
● It stores data locally, which means browsers and
websites will load faster because access elements
such as homepage images have previously been
downloaded.
Architecture

Without Cache With Cache

Type of Caching

In-Memory Cache

Question : Name all Indian

Caching is divided into two part as per the states
type of storage used.
1. In Memory Caching 1. Approach one for me:
Memorize all the names.
2. Distributed Caching
Distributed Cache

2. Approach : Write down on

my hand and read it.
 In Memory Cache

● In memory cache is when we use the application itself

as a storage service for cache.

● Usage in memory HashMaps or other data structures

to implement
● It is very fast as there are no external network calls.
Why not in memory cache?

● Complex to code and maintain

● Application goes down, cache goes down.
● As it usage RAM, the storage is very expensive.
● Requires notifying instances once a change occurs and need to update each instance separately
● Need to load mandatory cache on start of each server
● As each instance has its own cache there might be data inconsistency in the scenario where one instance
having different or outdated data than others.
Distributed Cache

● A cache shared by multiple app servers,

maintained as an external service to the app
servers that access it.
● Much simpler to maintain
● Even if application goes down cache will be
maintained as external data storage is used.
● Very reliable
● Only one place to update
● Assures data consistency
● Examples : Redis, ElastiCache, Hazelcast …
Redis Redis stands for Remote Dictionary
Server

● Redis is a fast, open source, in-memory,

key-value data store.

● Redis is written in C.

● It supports different kinds of abstract data

structures, such as strings, lists, maps, sets,
sorted sets,
HyperLogLogs, bitmaps, streams, and spatial
indices.
Which makes it a better choice than other
caching applications.

● It also provide clustering, replication, range

queries, Eviction Policies.
Redis topologies
1. Redis Stand Alone

2. Redis Master-Slave

3. Redis Sentinel

4. Redis Cluster
1. Redis stand alone

● A Single Redis node working as the cache.

● Easy to implement and maintain
● Fast, as only one network call, no addition proxy.
● Consistent, as only one node data will always be
consistent
● Single Point of Failure, node goes down entire cache
goes down.
● No failover, If node goes down, someone will have to
manually restart the redis server.
2. Redis Master-Slave

● In this topology multiple slave nodes are attached to

one Master slave.
● Data will be updated on this master and then the
master will push those changes to its replicas.
● Slaves can talk to the master only, and are only read
only nodes.
● In case of node failure:
○ If slave fails : it works as it is. as all calls goes to
master only.
○ If master fails : data won’t be lost as we have
replicas but we will have to manually
reconfigure and set any of the slave as master
● Client will also have to update the write node IP as
the master is changes.
3. Redis sentinel
● Redis Sentinel is a high availability solution provided by Redis,
which works on Master-Slave architecture.

● In this topology we have one sentinel node attached to each

redis node. The Sentinel decide the master node or we can
specify as well in first iteration.

● We can additionally specify to redirect all reads to slaves

which decrease load on Master and increase efficiency.

● Availability is extremely high.

● Automatic failover. If a master node fails, Sentinel starts the
failover process where a replica is promoted to master, the
other additional replicas are reconfigured to use the new
master.
● Monitoring. Sentinel constantly checks if your master and
replica instances are working as expected.
● Notification. Sentinel can notify the system administrator, or
other computer programs, via an API, that something is wrong
with one of the monitored Redis instances.
● If we have a large number of keys, it doesn’t solve horizontal
scaling.
4. Redis cluster

● Redis cluster solves the horizontal scaling problem,

where data is sharded among multiple master Nodes.

● Replication : Each master can have as many slaves as

required upto 1000. Min 1 slave is required.

● Must have at least 6 Redis nodes – 3 for masters and

three for slaves.

● There are two major benefits.

○ If master fails, 66% of application will face 0 down
time.
○ After the Failover is completed the slave will
become master and it will start working as it was.
4. Redis cluster
● Main purpose of the Redis cluster is to equally/uniformly
distribute your data load by sharding
● Redis Cluster does not use consistent hashing, but a
different form of sharding where every key is conceptually
part of what is called as hash slot
● There are 16384 hash slots in Redis Cluster, Every node in a
Redis Cluster is responsible for a subset of the hash slots,
so, for example, you may have a cluster with 3 nodes,
where:
Node A contains hash slots from 0 to 5500, Node B
contains hash slots from 5501 to 11000, Node C contains
hash slots from 11001 to 16383
● This allows us to add and remove nodes in the cluster
easily. For example, if we want to add a new node D, we
need to move some hash slot from nodes A, B, C to D
● Redis cluster supports the master-slave structure, you can
create slaves A1,B1, C2 along with master A, B, C when
creating a cluster, so when master B goes down slave B1
gets promoted as master
 Redis Sentinel vs Cluster
Redis Sentinel is a good option for a smaller implementation with high availability concerns. On the other hand, Redis Cluster is a clustering solution
that handles the scanning for larger implementations.

Metric Redis Sentinel Redis Cluster

Architecture Better visibility to the users. Data sharding, maintains happens internally

Availability Availability is high here as we the system can stand totalNumberOfNodes-1 Even if we have 6 nodes in our system, if one master slave pair goes
failures. down we will lose a lot of data. To Erase it we will require more number
Additionally as we have Sentinel nodes to monitor the nodes and decision of slave nodes per master. Increase cost.
making, makes it more available.

Cost Cheaper as you can start with only 2 nodes. Expensive as it requires min 6 machines.

Scalability We can vertically scale by increasing the size of master node but we have Scalability is great here as we can add as many master as we want.
a certain limitation to that as well.

Data Sharding No data sharding, as each node (master or slave) have the complete data Data is sharded among number of partitions created or number
copy. of master slave groups.