14
Put Data to Work. Simply. An Introduction to Crate Johannes Moser, Head of Integrations

Basic Introduction to Crate @ ViennaDB Meetup

Embed Size (px)

Citation preview

Put Data to Work. Simply.An Introduction to CrateJohannes Moser, Head of Integrations

About Me

• Loves working with people and code• Head of Integrations at @CrateIO

[email protected]@joemoeAT

Put DATA TO WORK. SIMPLY

Crate is a fully sharded, distributed, horizontally scaling SQL database with realtime search and aggregation to build worry-free applications and services at scale.

super simple to use & operateplays well with others delivers fast results

Crate is...

• distributed standard SQL• shared nothing architecture• auto-sharding/partioning/replication • extremely simple to install/operate• all data (relational, json docs, blobs) • realtime search & aggregations• horizontal scaling, elastic, resilient• dynamic schemas, agile dev• eventual consistent, atomic on doc level• balanced memory-disk usage• ideal for containers

Crate - Stateful Container enabled through masterless shared nothing architecture

c r a t e

APP CONTAINERS

LOCALSTORAGE

c r a t e

APP CONTAINERS

LOCALSTORAGE

c r a t e

APP CONTAINERS

LOCALSTORAGE

c r a t e

APP CONTAINERS

LOCALSTORAGE

C R A T E C L U S T E R

C R A T E C o n t a i n e r C L U S T E R C r a t e i s l i k e a n e t h e r - a n   o m n i p r e s e n t , p e r s i s t e n t l a y e r f o r y o u r d a t a

Crate Clusters running

right now (ca 400)

Competition

• MySQL• lacks scalability, fault-tolerance, search

• Hadoop World• it´s never realtime and not hot data

• Cassandra• rather inflexible, complex indexing, fat stack

• Elasticsearch• no SQL, not a DB, no compute of result

• MemSQL• memory only really, closed source

• MongoDB• slow aggregations, lack of search, tough scaling

• traditional, proprietary SQL servers• lack of scaling, complex and $$$

Customer “Skyhigh Networks”

• Security Events (times series)• Realtime Dashboard• side-by-side to Hadoop cluster• 500 customers running on Crate• better spread of data & load &

resilience• 3,2 BN inserts/updates per day• 48.000 concurrent TCP connections• about 60 nodes on AWS• 100% drop-in for MySQL cluster

-75% servers, but 20x performance

Use Cases for crate

Well Suited

• Real time analytics and business intelligence, Dashboards

• Internet of Things backend• High volume, semistructured/

dynamic data• Operational database for web

applications at scale • Combination of operational DB and

analytics DB

Not Well Suited

• Systems that require strong transactional consistency

• Strong, complex relational data(e.g insurances…)

DEMO

Crate Components (     Crate       Elasticsearch ,      other Open Source) 

Crate OPEN SOURCE STACK

Benefits FOR

developerS

• run faster !• use SQL and ORMs• migrate easily from existing solutions• native search & text processing• combine all data (relational, documents,

blobs) in one store• schema changes any time without penalty• times series and geospatial support• run Crate on your notebook• deploy app & database together (with

Docker)Ruby

Benefits OPS

• horizontally scalable, elastic• high availability out of the box• ease of use, simple operation• massively parallel read & write• incremental backup & restore• on prem & cloud integrations &

containers• manage TB-clusters for realtime queries• available for Linux, Mac, Windows• run in containers:

Docker, CoreOS, Tutum, Containership

THANK YOUJohannes Moser, Head of Integrations@joemoeAT, [email protected]