Slide 1
Slide 1 text
Paul Adams | Engineering Manager
testing crate with
docker engine 1.12
Slide 2
Slide 2 text
what is crate?
HIGHLY AVAILABLE - DEPLOY ANYWHERE
DYNAMIC HYPERGROWTH DATASETS
INCL TIMESERIES & GEOSPATIAL
Big Data Realtime SQL database
“TRANSLYTICAL PROCESSING”
HYPBRID OF TRANSACTIONAL AND ANALYTICS
REALTIME SEARCH & AGGREGATION
Slide 3
Slide 3 text
Crate is...
• distributed SQL via HTTP & TCP
• on NoSQL architecture scale out
• realtime search & aggregations
• all data: relational & JSON & blobs
• extremely simple to install/operate
• highly available
• auto-sharding/replication
• dynamic schema
• timeseries (partioning) / geospatial
• plug-in infrastructure
• consistent on row level (not ACID)
• super performant (in-memory style)
• Clients available for most languages
• runs on prem,AWS,GCE,Azure,Mesos,Docker…
Slide 4
Slide 4 text
CONTAINERS love
Crate
crate
APP CONTAINERS
LOCAL
STORAGE
c r at e
APP CONTAINERS
LOCAL
STORAGE
c r at e
APP CONTAINERS
LOCAL
STORAGE
CR ATE CLUSTE R
c r at e
APP CONTAINERS
LOCAL
STORAGE
Slide 5
Slide 5 text
IoT, Web, Mobile
SQL API
BI, Visualization,
Dashboards
AWS, GCE, Azure,
on premise, local
Direct installer,
Docker, Mesos,
Kubernetes, CoreOS
Plays well with others
Slide 6
Slide 6 text
Come and play!
• github.com/crate
• crate-community @ slack
• #crate @ freenode
• Berlin Crate Meetup
• SF Crate Meetup
• @crateio
Slide 7
Slide 7 text
time for the
risky demo
• 8 x Type 1 machines at packet.net
• 8 cores
• 32GB RAM
• 2 x 1Gb bonded network ports
• Orchestration courtesy of Docker
Engine 1.12