Infinispan , transactional key- value DataGrid and NoSQL database

Infinispan, transactional key-value DataGrid and NoSQL database11. April 2013 Alexander Petrov

Alexander Petrov

• Sr. Consultant at Inmeta Consulting

• Current project: Skattetaten Grid POC

• Previous projects involving grid technologies:• Mattilsynet food authority system.• FrameSolution BPM framework used in Lovisa National Court

Authority(Norway), Mattilsynet Food Authority

• Other noteworthy projects• Coca Cola Basis ERP system – Coca Cola Bottler factories• mPower Mobilitec 300 million subscribers worldwide, and delivers over

500,000 pieces of content every day.

Usage scenarios

• Big data, Databases are slow. Memory is FAST!• Provides huge computing power.• Tax calculation • Financial organizations• Government organizations use it for communication and

data sharing between the different departments.• Scientific computations• MMORPG games

Agenda

• General terminology relevant to Distributed Caching

• Challenges related to introducing distributed caching to existing system

• Metrics and tuning

Distributed Caching - Concepts

• Cache JSR – 107• Java Data Grid JSR - 347• In memory Data Grid• Cluster• Distribution• Node – a member of a cluster• Transaction awareness• Colocation• Map / Reduce• Consistency

Real World Use Case

Typical J2EE backend

Data access

• Transaction scope• Locking\deadlocking• Flushing policies• Mixing the technologystack.• Performance

Legacy Cache

Our end goal

• Wow we did it!

Summary

• Our Custom cache is super fast, but its cache hit ratio is rather low.

• Our custom cache has a tendency of getting dirty as the updates to the shared data can not be propagated. At the same time the separation of the data regions is not full.

• Marshaling is a rather slow and heavy process.

• We are facing a technological cocktail and we need to keep integrity.

Replication

• Write through• Write Behind• Replication Queue

Invalidation

Distribution

More terminology

• Eviction• Least Recently Used• First In First Out• LIRS• Custom

• Expiration

• Invalidation

Caching topologies – Mirrored Cache

• Ref. Data vs Transactional• Reference data: Good.Max 30000 reads/sec 1k size• Transactional data: Good.Max 25000 writes/sec 1k size

.

Caching topologies – Replica Cache

• Reference data: Good.30000 reads/sec per server.Grow linearly by adding servers.

• Transactional data: Not sogood. Max 20000writes/second. Drops if you add 3rd server to2500.

Caching topologies – Partitioned Cache

• Ref. Data vs Transactional• Reference data: Good.Max 30000 reads/sec 1k size• Transactional data: Good.Max 25000 writes/sec 1k size

Caching topologies - Partitioned Replica

• Reference data(1kb):Good. 30000 reads/sec per server. Grow linearly by adding servers.

• Transactional data(1kb):Good. 20000 writes/sec per server.Grow linearly by adding servers.

How to define our topology

• What is the size of our cluster? Reads vs. Writes• Communication inside our grid

• UDP,TCP• Synchronous vs. Asynchronous.• What about the transaction isolation?

• Repeatable Reads vs. Read Committed• What is the nature of our application?• Read intensive data

• CMS systems • Write Intensive Data

• Document Management System

Level 1 Cache / Near Cache

• Level1 cache is Supported only for Distribution mode• Level 1 cache mighthave a performance Impact in certain systems

Cache stores and loaders

• Passivation

• Activation

• Hibernate

Transactions, Isolation and Locking

• Long running transactions need to be avoided.

• What is a long running transaction? How long is actually long.

• Read Committed vs Repeatable Reads

Classic Deadlock situation

begin Update(A) Update(B) Update(C) Update(B)

Begin Update(C) Update(B) Release(A) Lock(A)

TX1 (Wants update A,B,C)

TX2 (Wants to update C,B,A)

C is locked by TX2

A is locked by TX1

Repeatable Read

begin get(k) - - Get(k)

Begin Get(k) put(k, v2) commit

What is returned??

TX1

TX2

Cache statistics

Remoting statistics

Locking statistics

Marshaling data

• Java serialization

• Java externalization

• Impact on performance

• Generic domain.

Real World Use Case

Data access

• Transaction scope• Locking\deadlocking• Flushing policies• Mixing the technologystack.• Performance

Our end goal

• Wow we did it!

The End

• Thank you for your attention

Used sources

http://www.alachisoft.com/ncache/caching-topology.htmlhttp://www.infoq.com/news/2011/10/java-data-gridhttps://github.com/datagrids/spec/wikihttp://www.jboss.org/infinispan/documentationhttp://code.google.com/p/thrift-protobuf-compare/wiki/Benchmarking