Distributed messaging through Kafka

New age Distributed Messaging
Kafka & Concepts explored !!
Dileep Varma Kalidindi
Nov 2014

Who Am I ?
4/5/2016 Confidential 2
Name: Dileep Varma Kalidindi
Status: Senior Engineer @Responsys (since Apr’14), Circles Team.
Fascination: Problem Solving , Distributed & BigData churning systems.
Past: 8+yrs with VeriSign, Informatica Labs, NTT Data.
Hobbies: Jumping (Water & Air)

What is brewing today ?
 Responsys Technology Road Map.
 Data off the limits - Handling & Processing BigData
 Scope for New Age capabilities (in distributed msg’ng) – Architecture peek through
 Existing System bottlenecks & shortfalls
 Rethinking from fundamentals – Distributed Commit Log
 Kafka Messaging – Concept, Architecture, API & Demo
 Kafka Internals – ZooKeeper in depth, Atomic broadcast & Quorum
 Performance & feature comparisons – Traditional vs New Age

Are we good ?

Data off the limits – Handling larger Data sets
 Kafka on Responsys technology Road map - Antonio
 Data evolution from Traditional to BigData
 Characterized by Volume, Variety, Velocity, Variability, Veracity & Complexity
 Volume -> Quantity of data. Storage & Processing (Hadoop, NoSQL)
 Variety -> Diversity of data sets, OLTP, OLAP (NoSQL, NewSQL)
 Velocity -> Speed of data handling in real time (Kafka, Storm, Flume)
 Deeper market penetration implicitly transforms Data
 Our focus is on Velocity
 Need of the hour is Systems to handle – BigData Technologies

BigData Technologies – MindMap view

7
Uber
Application
Database
UI PUB WS CN BounceIS
LA JMS EC SPAM ETLAB
EventDB
CustDB
ReportDBSysAdmDB
Data
Warehouse
AuditDB
UsageDB
EMD CL PD
ICR
Content
IDDP
Short URL
SUL DIS
SMS PGPUSH
SMSL
Identifying Scope – Architecture Peek-in
REAL TIME PROCESSING

Is a there problem with my current System ?
 Existing systems are good (IBM MQ) in traditional sense.
 Delivery guarantee is good for Emails, what for events (PubWeb, Bounce, AB) ?
 Focus on throughput. Existing brokers have limitations.
 Scaling and Replication, cost of Cluster maintenance in existing MQ.
 Dynamic rebalancing of Brokers, Consumers

Rethink from Fundamentals
LOGS

Log’s – fundamental System blocks
• Log (as a foundation) :
 Append-only, totally-ordered sequence of records ordered by
time.
 Unique –sequential log entry (Clock Decoupled time stamp)
 Deterministic
• Logging (as a core process) :
• IS Machine readable logging
Ex: Write ahead logs, Commit logs &Transaction logs
• IS NOT Application logging (Human readable)
Ex: Log4j, slf4j etc..
• Backbone of Distributed Messaging, Databases, NoSQL, Key-
Value stores, replication, Hadoop, Version Control…
• Logs for Data Integration, Real time processing & System
building.

Log’s – solving Problems
• Logs are not new in Databases !!
 Started with IBM SystemR
 Physical logging – Values of rows changed, Logical logging – SQL Queries
 Logs implementations – ACID to Replication (Goldengate)
• State Machine Replication Principle
2 identical, deterministic process -> begin with same state, gets same inputs in order, produce same output and
ends in same state
• In Distributed Systems they Solve core problems
Ordering changes Distributing data
• Processing and replication
Active – Passive
Active - Active

Log’s – driving Architecture
• Log-structured data flow
 Cache system
 Asynchronous Production & Consumption
• Kafka Log Centric approach:
 Not a Database, Log file collection, Typical messaging system
• Event driven architecture:
 Kafka – event driven, Multi-subscriber system (Topic)
 Example – which performs multiple ops on one event job

Logs in ACTION
APACHE KAFKA

Kafka
4/5/2016
Confidential 14
 Introducing Kafka
“Should I wake-up now? ..why ?  “
 Kafka Core Concepts
Topics, partitions, replicas, producers, consumers, brokers
 Operating Kafka
Architecture, deploying, monitoring, P&S tuning

Introducing Kafka
4/5/2016
Confidential 15
http://kafka.apache.org/
Originated at LinkedIn, open sourced in early 2011
Implemented in Scala, some Java
9 core committers, plus ~ 20 contributors
Kafka is a distributed, partitioned, replicated commit log
service. A uniquely designed pub-sub messaging system
Designed for,
 High throughput to support high volume event feeds.
 Support real-time processing of these feeds to create new, derived feeds.
 low-latency delivery to handle traditional messaging use cases.
 Guarantee fault-tolerance

Kafka in Real business
4/5/2016
Confidential 16

Kafka is Amazingly fast – How ?
4/5/2016 17Confidential
• “Up to 2 million writes/sec on 3 cheap machines”
• Using 3 producers on 3 different machines, 3x async replication
• Only 1 producer/machine because NIC already saturated
• Sustained throughput as stored data grows
• Slightly different test config than 2M writes/sec above.

Kafka is Amazingly fast – Why ?
• Fast writes:
• While Kafka persists all data to disk, essentially all writes go to the
page cache of OS, i.e. RAM.
• Cf. hardware specs and OS tuning (we cover this later)
• Fast reads:
• Very efficient to transfer data from page cache to a network socket
• Linux: sendfile() system call
• Combination of the two = fast Kafka!
• Example (Operations): On a Kafka cluster where the consumers are mostly caught
up you will see no read activity on the disks as they will be serving data entirely
from cache.

Kafka Core Concepts - A first look
• The who is who
• Producers write data to brokers.
• Consumers read data from brokers.
• All this is distributed.
• The data
• Data is stored in topics.
• Topics are split into partitions, which are replicated

Kafka Concepts - Topics
• Topic: feed name to which messages are published
• Example: “pubweb.event.2”

Kafka Concepts - Topics

Kafka Concepts -Creating a Topic
• Creating a topic
• CLI
• API
https://github.com/miguno/kafka-storm-
starter/blob/develop/src/main/scala/com/miguno/kafkastorm/storm/KafkaStorm
Demo.scala
• Auto-create via auto.create.topics.enable = true
• Modifying a topic
- Add partitions
- Add configs
- Remove Configs
- Deleting topics
$ kafka-topics.sh --zookeeper zookeeper1:2181 --create --topic zerg.hydra
--partitions 3 --replication-factor 2
--config x=y

Kafka Concepts - Partitions
• A topic consists of partitions
• Partition: ordered + immutable sequence of messages
that is continually appended to
• Partitions of a topic are Configurable

Kafka Concepts - Partition Offset
• Offset: messages in the partitions are each assigned a unique (per
partition) and sequential id called the offset
• Consumers track their pointers via (offset, partition, topic) tuples
Consumer group C1

Kafka Concepts - Partition Replica’s
• Replicas: “backups” of a partition
• They exist solely to prevent data loss.
• Replicas are never read from, never written to.
• They do NOT help to increase producer or consumer parallelism!

Topics vs Partitions vs Replica’s

Kafka Concepts - Topic inspection
• --describe the topic
• Leader: brokerID of the currently elected leader broker
• Replica ID’s = broker ID’s
• ISR = “in-sync replica”, replicas that are in sync with the leader
• In this example:
• Broker 0 is leader for partition 1.
• Broker 1 is leader for partitions 0 and 2.
• All replicas are in-sync with their respective leader partitions.
$ kafka-topics.sh --zookeeper zookeeper1:2181 --describe --topic zerg.hydra
Topic:zerg2.hydra PartitionCount:3 ReplicationFactor:2 Configs:
Topic: zerg2.hydra Partition: 0 Leader: 1 Replicas: 1,0 Isr: 1,0

Kafka Concepts - Consumers & Producers
df

Kafka Concepts - Producer
df
• Code
• Start Producer

Kafka Concepts - Consumers
df
• Code
• Start Consumer
• Multithreaded Consumer for multiple
partitions

Kafka Core Concepts - Recap
• The who is who
• Producers write data to brokers.
• Consumers read data from brokers.
• All this is distributed.
• The data
• Data is stored in topics.
• Topics are split into partitions, which are replicated

Monitoring & Testing

Kafka – Monitoring and Testing
• JMX Enabled
• System tools
• Describe
• Quantified Offset Monitor
• Monitoring DEMO

Empowering Kafka

Apache ZooKeeper
4/5/2016
Confidential 35
 Apache Kafka uses ZooKeeper to detect crashes, implement topic
discovery, and maintain production & consumption state for topics.
 High-performance coordination service for distributed applications.
 SoC – Separates Coordination overhead from Application logic.
 Centralized service for naming (registry), configuration
management, synchronization, and group membership services.
 Zookeeper is backbone for Hbase, Solr, Facebook messaging apps &
many more distributed apps.
 Simple, Replicated, Ordered and Fast

Zookeeper- Internals
4/5/2016
Confidential 36
 Znodes
 Persistent – exists till deleted
 Ephemeral - session scope
 Reads by all Nodes and Writes through Leaders
 Data is stored as byte array
 Allows Watches and notifications
 Ensemble – a group of Servers available to service
 Quorum determined leader selection

ZooKeeper – Guarantees
4/5/2016
Confidential 37
• Follows principles of ATOMIC broadcast
 Sequential Consistency – Updates are applied in order
 Atomicity – Updates either succeed or fail
 Single system image – Same view of service regardless of ZK server
 Reliability – Persistence of updates
 Timeliness – System is guaranteed to be up-to-date within time bound
• In Summary - Zookeeper { Leader Activation + Message delivery }

Kafka Performance

Kafka performance – Producer tests
(LinkedIn benchmark test)
4/5/2016
Confidential 39
• HW Set-up with 2 linux nodes
• Each with 8 2 GHZ cores (8 Cores/Mac ~ 16 GHZ processing)
• 16 GB of RAM, 6 disks with RAID 10 and 1GB network connection.
• Producer test
• Single producer ~ 10 million msgs each of 200bytes
• Kafka msg batch 1 and 50. Other MQ’s no batching
• X-axis – Msg sent to broker, Y-axis – Producer throughput
• Why is Producer fast
• No ACK
• Batching
• Kafka storage format

Kafka performance – Consumer tests
(LinkedIn benchmark test)
4/5/2016
Confidential 40
• HW Set-up with 2 linux nodes
• Each with 8 2 GHZ cores (8 Cores/Mac ~ 16 GHZ processing)
• 16 GB of RAM, 6 disks with RAID 10 and 1GB network connection.
• Consumer test
• Single consumer retrives 10 million msgs each of 200bytes
• Each pull request for 1000 msgs (200kb)
• X-axis – Msg consumed from broker, Y-axis – consumer throughput
• Why is Producer fast
• No Delivery state storage
• Kafka storage format
(less data transmitted)

Summary, Conclusions
&
References

Summary – quick Recap
4/5/2016
Confidential 42
 Importance Handling & Processing BigData
 Scope for introduction in Responsys Architecture
 Existing System bottlenecks & shortfalls
 Distributed Commit Log
 Kafka Messaging
 Kafka Internals – ZooKeeper
 Performance & feature comparisons – Traditional vs New Age

Conclusion – Open ended
4/5/2016
Confidential 43
• Limitation is on Data – not on Systems
• No need for complete revamp
• Choice of Right systems at right time is the recipe.
References
1. https://kafka.apache.org/
2. http://zookeeper.apache.org/
3. http://engineering.linkedin.com/kafka/benchmarking-apache-kafka-2-million-writes-second-
three-cheap-machines
4. http://engineering.linkedin.com/distributed-systems/log-what-every-software-engineer-
should-know-about-real-time-datas-unifying

THANK YOU

Distributed messaging through Kafka

More Related Content

What's hot

Similar to Distributed messaging through Kafka

Recently uploaded

Distributed messaging through Kafka