AWS ElastiCache Deep Dive: Best Practices and Usage Patterns

© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
AWS re:INVENT
Amazon ElastiCache Deep Dive:
Best Practices and Usage Patterns
M i c h a e l L a b i b , N o S Q L S p e c i a l i s t S A , A W S

What to expect from this session
• Amazon ElastiCache overview
• Scaling your cluster with online re-sharding
• Amazon ElastiCache security and encryption
• Amazon ElastiCache usage patterns
• Best practices

Amazon
ElastiCache
µs is the new ms

In-memory key-value store supporting
• Redis 3.2.10
• Memcached 1.4.34
High-performance
Fully managed; zero admin
Highly available and reliable
Hardened by Amazon
Amazon
ElastiCache

Request rate
High Low
Latency
Low High
Structure
Low
High
Data volume
Low High
Amazon
RDS
Amazon S3
Amazon
Glacier
Amazon
CloudSearch and
Amazon Elasticsearch
Service
Amazon
DynamoDB
Amazon
ElastiCache
and Amazon
DynamoDB
Accelerator
(DAX)
HDFS

Powerful
~200 commands + Lua scripting
In-memory data structure server
Utility data structures
Strings, lists, hashes, sets, sorted sets,
bitmaps, and HyperLogLogs
Simple
Atomic operations
supports transactions
Ridiculously fast!
<1 ms latency for most commands
Highly available
replication
Persistence
Open source
Redis overview

But wait, there’s more!
Run Lua scripts Geospatial queries Pub/sub

SMEMBERS features
REDIS:6379>
Amazon
ElastiCache
1) “Easy to deploy & monitor”
AWS
Config
Amazon
CloudWatch
AWS
CloudTrail
AWS
CloudFormation
AWS
Management
Console
AWS CLI
and SDKs
alarm
REDIS:6379>
hget feature:details “deploy-monitor”
Amazon
SNS
Email
Notification
AWS
Lambda

SMEMBERS features
REDIS:6379>
REDIS:6379>
hget feature:details “enhancements”
2) “Enhanced Redis Engine”
Optimized Swap Memory
•Mitigate the risk of increased swap usage
during syncs and snapshots
Dynamic write throttling
•Improved output buffer management when
the node’s memory is close to being
exhausted
Smoother failovers
•Clusters recover faster as replicas avoid
flushing their data to do a full re-sync with
the primary
Amazon
ElastiCache

SMEMBERS ec-team:open-source:contributorsREDIS:6379>
https://raw.githubusercontent.com/antirez/redis/2.8/00-RELEASENOTES
(from https://github.com/antirez/redis/issues/1814)
1) “Kevin McGehee”
2) “Qu Chen”
3) “Rajib Dugar”
hgetall ec-team:open-source:contributionsREDIS:6379>
1) PSYNC2 (Redis 4.0)
3) BGSAVE (Redis 3.2)
5) MIGRATE (Redis 3.0)
7) MASTER TIMEOUT (Redis 2.8)
9) INCREASE 2Billion+ KEYS in a DATASET (Redis 2.8)
2)
4)
6)
8)
10)
Amazon
ElastiCache
11) also: 4114, 4250, 3926, 3899

Slot 0-5461
Cluster Mode Disabled
Keyspace
Slot 10923-16383
I Primary 0–5 Replicas
Cluster Mode Enabled
Primary Endpoint
1–15 Primaries/Shards
Slot 0
Slot 5462-10922
Slot 16383
Keyspace
0–5 Replicas
Configuration Endpoint
Slot 1 …
Vertically Scaled
Horizontally Scaled
Max Storage 407 GiB
Max Storage 6+ TiB
Redis topologies

Feature Enabled Disabled
Failover 15–30 sec
(Non-DNS)
~1.5 min
(DNS-based)
Failover risk • Writes affected—partial dataset (less risk with
more partitions)
• Reads available
• Writes affected on entire dataset
• Reads available
Performance Scales with cluster size
(90 nodes—15 primaries + 0–5 replicas per shard)
6 nodes (1 primary + 0–5 replicas)
Max connections • Primaries (65,000 x 15 = 975,000)
• Replicas (65,000 x 75 = 4,875,000)
• Primary: 65,000
• Replicas: (65,000 x 5 = 325,000)
Storage 6+ TiB 407 GB
Cost
Example: Assume
workload needs 175 GB
Smaller nodes but more $$
9 x cache.r3.xlarge ($0.455hr) = $4.095 hr 255.6 GB
Larger nodes less $
1 X cache.r3.8xlarge = $3.640, 237 GB
Redis cluster-mode enabled vs. disabled

Closer look at cluster-mode enabled

S5
S1
S2
S4 S3
Client
 16384 hash slots per cluster
 Slot for a key is CRC16(key) mod 16384
 Slots are distributed across the cluster into shards
 Developers must use a Redis cluster aware client
 Clients are redirected to the correct shard
 Smart clients store a map
Shard S1 = slots 0–3276
Shard S5 = slots 13107–16383
Redis cluster: automatic client-side sharding

Availability Zone A
slots 0–5454 slots 5455–10909
Redis Cluster
Redis cluster—architecture
slots 10910–16363
Availability Zone B Availability Zone C
slots 5455–10909
slots 5455–10909slots 0–5454 slots 0–5454
slots 10910–16363
slots 10910–16363
Redis cluster—multi-AZ
A cluster consists of 1 to 15 shards
Example: 3-shard cluster,
2 read replicas

Availability Zone A
slots 0–5454
Redis Cluster
slots 10910–16363
slots 5455–10909
slots 5455–10909slots 0–5454 slots 0–5454
slots 10910–16363
Shard
ReplicaReplicaPrimary
Each shard has a primary node
and up to five replica nodes
slots 5455–10909
slots 10910–16363

Availability Zone A
slots 0–5454 slots 5455–10909
Redis Cluster
slots 10910–16363
slots 5455–10909
slots 5455–10909
Shard
ReplicaReplica Primary
slots 0–5454 slots 0–5454
slots 10910–16363
slots 10910–16363

Availability Zone A
slots 0–5454
Redis Cluster
slots 10910–16363
slots 10910–16363
slots 10910–16363
Shard
Replica PrimaryReplica
slots 5455–10909 slots 0–5454
slots 5455–10909
slots 0–5454 slots 5455–10909

Availability Zone A
slots 0–5454 slots 5455–10909
Redis Cluster
slots 10910–16363
slots 5455–10909 slots 5455–10909slots 0–5454 slots 0–5454
slots 10910–16363 slots 10910–16363
Scenario 1: Single primary failure

Availability Zone A
slots 0–5454 slots 5455–10909
Redis Cluster
Scenario 1: Single primary failure
slots 10910–16363
slots 10910–16363
Mitigation:
1. Automatic failure detection and replica promotion (~15–30 s)
2. Repair failed node
slots 10910–16363

Availability Zone A
slots 0–5454 slots 5455–10909
Redis Cluster
Scenario 2: Majority of primaries fail
slots 10910–16363
slots 10910–16363slots 10910–16363

Availability Zone A
slots 0–5454 slots 5455–10909
Redis Cluster
slots 10910–16363
Mitigation: Redis enhancements on ElastiCache
• Automatic failure detection and replica promotion
• Repair failed nodes
slots 10910–16363slots 10910–16363
Scenario 2: Majority of primaries fail

aws elasticache create-snapshot --replication-group-id redisclusterID --snapshot-name snameStep 1
aws elasticache copy-snapshot --source-snapshot-name sname --target-snapshot-name sname
--target-bucket s3ucketname
Step 2
Step 3 aws elasticache create-replication-group --replication-group-id NewRedisClusterID … --snapshot-arns
arn:aws:s3:::bucketname/redisbackup-0001.rdb, etc.
Step 4 Once the new cluster is up, update your app with new Amazon ElastiCache endpoint, then terminate old cluster
3 Shards 5 Shards
Downtime
New writes
not in
snapshot
rdb
Pro tip: DR strategy—enable
CRR on Amazon S3 bucket
triggering AWS Lambda function
to hydrate destination cluster
Resizing via backup and restore

Zero-downtime Online Re-sharding

0-5461
Shard 1 Shard 2 Shard 3
5462--10922 10923-16383
aws elasticache modify-replication-group-shard-configuration --replication-group-id rep-group-id
--apply-immediately --node-group-count 5
Simple API
Scale In || Out
Online Re-Sharding—zero downtime

0-5461
reads/writes
Shard 4 Shard 5
5462--10922 10923-163830-2909,
5095-5461
5462-5783,
6876-9830
10923-14199
2910-5094,
9831--10922
No application interruption
Uniform slot distribution across shards
5784-6875,
14200-16383
Online Re-Sharding—zero downtime: Scale Out

0-5461
reads/writes
Shard 4 Shard 5
5462--10922 10923-16383
Uniform slot distribution across shards
No Application Interruption
Online Re-Sharding—zero downtime: Scale In

AWS Lambda
3 Shards
…
var params = {
ApplyImmediately: true,
NodeGroupCount: 5,
ReplicationGroupId: ‘rep-group-id’,
… }
elasticache.modifyReplicationGroupShardConfiguration(params, function(err, data) {
if (err) console.log(err, err.stack);
else console.log(data);
}); …
5 Shards
MEMORY
HIGH!
Amazon
CloudWatch
Cluster Resized
AWS SNS
Online Re-Sharding—CW alarm triggered

reads/
writes
reads
AZ1
AZ2
reads
search
reads
search
clients
c a c h e c l u s t e r
r e l a t i o n a l
d a t a
Healthy

reads/
writes
reads
AZ1
AZ2
reads
search
reads
search
clients
r e l a t i o n a l
d a t a
Heavy
pressure

reads/
writes
reads
AZ1
AZ2
reads
search
reads
search
clients
r e l a t i o n a l
d a t a
Healthy—
auto scaled
out

Amazon ElastiCache security

ElastiCache Redis reference architecture
Availability Zone B Availability Zone CAvailability Zone A
REDIS:6379> hget feature:details “ref-arch”

ElastiCache Redis Reference Architecture
Private SubnetPrivate Subnet Private Subnet

security group security group security group

ElastiCache Redis Cluster
Amazon S3
bucket
REDIS RDB
snapshot
Encryption at
REST

ElastiCache Redis Cluster
security group
Public Subnet
security group
Public Subnet Public Subnet
Amazon S3
bucket
REDIS RDB
snapshot
security group
Encryption In-Transit 3.2.6 Redis AUTH
Encryption at
REST

Encryption
• In-Transit: encrypt all communications between clients
and Redis server as well as between nodes
• At-Rest: encrypt backups on disk and in Amazon S3
• Fully managed: setup via API or console, automatic
issuance and renewal
Compliance
• HIPAA eligibility for ElastiCache for Redis
• Included in AWS Business Associate Addendum
• Redis 3.2.6
Amazon ElastiCache Encryption and Compliance

Common usage patterns

Session
management
Database caching APIs
(HTTP responses)
IOT
Streaming data
analytics
(Filtering/aggregation)
Pub/sub
Social media
(Sentiment analysis)
Standalone
database
(Metadata store)
Leaderboards
Usage patterns

Caching
Clients
Amazon
ElastiCache
Redis
Amazon
DynamoDB
Elastic Load
Balancing
Amazon
EC2
Amazon
RDS
write-through
reads/
writes
DDB streams
mysql.lambda_async
reads/
writes
Amazon
S3
reads/writes
Object data
Unstructured data
Relational data

Caching NoSQL
Amazon
EC2 reads/
writes
reads
MongoDB
Cluster
Cassandra
Cluster
 Smaller NoSQL DB clusters needed = lower costs
 Faster data retrieval = better performance
Elasticsearch
Cluster
Clients

Amazon
EC2
reads/
writes
Amazon
ElastiCache
Redis
reads
MongoDB
Cluster
DBObject doc = collection.findOne();
Cache serialized DBObject in Redis (good)
Cache rows in Redis hash (faster/more efficient)
Cassandra
Cluster
Amazon
ElastiCache
Redis
Amazon
EC2
reads/
writes
reads
ResultSet rs = session.execute(stmt);
Cache serialized ResultSet in Redis (good)
Cache rows in Redis hash (faster/more efficient)
 Smaller NoSQL DB clusters needed = lower costs
 Faster data retrieval = better performance
Caching NoSQL databases with Amazon ElastiCache

Amazon
Kinesis
Analytics
Amazon
Kinesis
Streams
Amazon
Kinesis
Streams
Amazon
ElastiCache
(Redis)
cleansed
stream
Streaming data enrichment/processing
Datasources
raw
stream
Subscribers
AWS Lambda function 1
Continual data
filtering/enrichment
Real-time
pub/sub
AWS Lambda function 2

Big data architectures using Redis
Amazon Kinesis
DataSources
AWS Lambda
Apache Storm
on EMR
Spark Streaming
on Amazon EMR
Amazon
Kinesis app
Amazon
EC2
AWS IoT
Amazon
ElastiCache
Collect
Store
Process
Amazon
S3
Apache Kafka
AWS
Lambda
Custom
app
Spark on
Amazon
EMR
Analyze

Rules Engine
Amazon
ElastiCache
Redis
AWS
Lambda
Direct integration
LambdaSNS SQS
S3 KinesisDDB
AWS
IoT devices
AWS
IoT
Sensor store
IoT powered by ElastiCache

Mobile apps powered by ElastiCache
Amazon API
Gateway
AWS
Lambda
Amazon
ElastiCache
Redis
GEOADD
GEORADIUS
Search points of interest
Update points of interest
https://aws.amazon.com/blogs/database/amazon-elasticache-utilizing-redis-geospatial-capabilities/
Amazon
DynamoDB
DDB streams
Amazon
EC2

Ad tech powered by ElastiCache
Clients
Advertisers
https://aws.amazon.com/caching/database-caching/
Ad network
Ad slot
Consumer
Ad slot
publishers
Ad placement
(websites/apps)
Amazon
ElastiCache
Redis
<40 ms
Clickstream
(shopping
events)
User visits
page
Publisher
places ad slot
for auction
Ad network
calls for bidsBidders respond
with bids
Winners bid
ad displayed

Chat apps powered by ElastiCache
https://aws.amazon.com/blogs/database/amazon-elasticache-utilizing-redis-geospatial-capabilities/
Clients
Chat apps
Application Load
Balancer
WebSockets
Amazon
ElastiCache
Redis
PubSub
Serverpersistent
connections
Elastic
Beanstalk
SUBSCRIBE chat_channel:114
PUBLISH chat_channel:114 "Hello all"
>> ["message", "chat_channel:114", "Hello all"]
UNSUBSCRIBE chat_channel:114

• Very popular for gaming apps that need
uniqueness and ordering
• Easy with Redis sorted sets
ZADD "leaderboard" 1201 "Gollum”
ZADD "leaderboard" 963 "Sauron"
ZADD "leaderboard" 1092 "Bilbo"
ZADD "leaderboard" 1383 "Frodo”
ZREVRANGE "leaderboard" 0 -1
1) "Frodo"
2) "Gollum"
3) "Bilbo"
4) "Sauron”
ZREVRANK "leaderboard" "Sauron"
(integer) 3
Gaming—real-time leaderboards

Ex: throttling requests to an API
uses Redis counters
ELB
Externally
facing API
Reference: http://redis.io/commands/INCR
FUNCTION LIMIT_API_CALL(APIaccesskey)
limit = HGET(APIaccesskey, “limit”)
time = CURRENT_UNIX_TIME()
keyname = APIaccesskey + ":” + time
count = GET(keyname)
IF current != NULL && count > limit THEN
ERROR ”API request limit exceeded"
ELSE
MULTI
INCR(keyname)
EXPIRE(keyname,10)
EXEC
PERFORM_API_CALL()
END
Rate limiting

Amazon ElastiCache: Best practices

Cluster sizing best practices
• Storage—clusters should have adequate memory
• Recommended: memory needed + 25% reserved memory (for Redis) + some room for growth
(optional 10%)
• Optimize using eviction policies and TTLs
• Scale up or out before reaching max-memory using CloudWatch alarms
• Use memory optimized nodes for cost effectiveness (R4 support)
• Performance—performance should not be compromised
• Benchmark operations using Redis Benchmark tool
• For more READIOPS—add replicas
• For more WRITEIOPS—add shards (scale out)
• For more network IO—use network optimized instances and scale out
• Use pipelining for bulk reads/writes
• Consider Big(O) time complexity for data structure commands
• Cluster isolation (apps sharing key space)—choose a strategy that works for your workload
• Identify what kind of isolation is needed based on the workload and environment
• Isolation: No Isolation $ | Isolation by Purpose $$ | Full Isolation $$$

Redis benchmark tool
Open source utility to benchmark performance
example: src/redis-benchmark -h r3-xlarge-perf.foio87.0001.use1.cache.amazonaws.com -p 6379 -n -150000 -d 100
Syntax:
redis-benchmark -h <host> -p <port> -c 50 -n 1000 -d 500 –q
-c <clients>—Specifies the number of parallel connections (default 50).
-n <requests>—Specifies the number of requests (default 1000000).
-d <size>—Specifies the data size of GET and SET values in bytes.
-t <test1,test2>—Comma-separated list of tests to perform.
-q—Quiet operation, displays only the result.

Redis max-memory policies
Select a max-memory policy based on your workload needs
• noeviction: return errors when the memory limit has been reached and the client is trying to execute
commands that might result in more memory to be used
• allkeys-lru: evict keys trying to remove the less recently used (LRU) keys first
• volatile-lru: evict keys trying to remove the less recently used (LRU) keys first, but only among keys
that have an expire set
• allkeys-random: evict random keys to make space for the new data added
• volatile-random: evict random keys to make space for the new data added, but only evict keys with an
expire set
• volatile-ttl: evict only keys with an expire set, and try to evict keys with a shorter time to live (TTL) first

Key ElastiCache CloudWatch metrics
• CPUUtilization
• Memcached—up to 90% ok
• Redis—divide by cores (ex: 90% / 4 = 22.5%)
• SwapUsage low
• CacheMisses/CacheHits Ratio low/stable
• Evictions near zero
• Exception: Russian-doll caching
• CurrConnections stable
• Setup alarms with CloudWatch metrics

ElastiCache modifiable parameters
• Maxclients: 65000 (unchangeable)
• Use connection pooling
• timeout—closes a connection after it has been idle for a given interval
• tcp-keepalive—detects dead peers given an interval
• Databases: 16 (default) for non-clustered mode
• Logical partition
• Reserved-memory: 25% (default)
• Recommended
 50% of maxmemory to use before 2.8.22
 25% after 2.8.22—ElastiCache
• Maxmemory-policy:
• The eviction policy for keys when maximum memory usage is reached
• Possible values: volatile-lru, allkeys-lru, volatile-random, allkeys-random, volatile-ttl,
noeviction

• Understand the frequency of change of underlying data
• Set appropriate TTLs on keys that match that frequency
• Choose appropriate eviction policies that are aligned with application requirements
• Isolate your cluster by purpose (for example, cache cluster, queue, standalone database, and so on)
• Maintain cache freshness with write-throughs
• Performance test and size your cluster appropriately
• Monitor Cache HIT/MISS ratio and alarm on poor performance
• Use failover API to test application resiliency
Caching tips

Thank you!

AWS ElastiCache Deep Dive: Best Practices and Usage Patterns

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to AWS ElastiCache Deep Dive: Best Practices and Usage Patterns

Similar to AWS ElastiCache Deep Dive: Best Practices and Usage Patterns (20)

More from Amazon Web Services

More from Amazon Web Services (20)

AWS ElastiCache Deep Dive: Best Practices and Usage Patterns