Felipe Mendes, Technical Director at ScyllaDB
Getting Started
with ScyllaDB
Agenda + Modern Scaling Challenges
+ What is ScyllaDB?
+ ScyllaDB Features
+ Getting Started & Next Steps
Yet Another
Database...?
Modern scaling challenges
Keep CapEx
& OpEx in check
Reduce complexity
Scale as the
data grows
Queries in
milliseconds
Leverage massive
amounts of data
Predictable, consistent
performance
Nestify – The Business Case for Investing in Web Performance Optimization
Time Matters
Database Evolution Over Time
Traditional RDBMS:
+ Structured data
+ Doesn’t scale
Document Store:
+ No structure
+ Limited scale
Key-Value Store:
+ Simple
+ Fast
+ Complex
Column Store:
+ Some structure
+ Best HA/DR
You probably use one or more of these for your organization :-)
Lower Node
Count
Efficient hardware usage
results in direct savings
What if we had it all in one…
Predictable,
Low Latencies
Consistent single-digit
millisecond p99 latencies
Less
Complexity
Self-optimizing, smaller
footprint, easy to use
Developer-friendly
APIs
Ease of migration and less
development cycles
What is ScyllaDB?
Predictable
Performance at
Scale
Scale Predictably
ScyllaDB and Memcached
We Compared ScyllaDB and Memcached and… We Lost?
Trillions of messages
60% fewer nodes
5msp99
“It’s a much more
efficient database —
we’re going from running
177 Cassandra nodes to
just 72 ScyllaDB nodes.”
2015 – Million
2016 – Billion
2019 – Trillion
Bo Ingram
How Discord Stores Billions and Trillions of Messages
+400 Gamechangers Leverage ScyllaDB
Seamless experiences
across content + devices
Connecting people all
around the globe
Corporate fleet
management
Real-time analytics 2,000,000 SKU -commerce
management
Video recommendation
management
Elastic Kafka Persistent
State Store at scale
Real time fraud detection
across 6M transactions/day
Uber scale, mission critical
chat & messaging app
Network security threat
detection
Power ~50M X1 DVRs with
billions of reqs/day
Precision healthcare via
Edison AI
Inventory hub for retail
operations
Property listings and
updates
Unified ML feature store
across the business
Cryptocurrency exchange
app
Geography-based
recommendations
Global operations- Avon,
Body Shop + more
Predictable performance for
on sale surges
GPS-based exercise
tracking
Serving dynamic live
streams at scale
Powering India's top
social media platform
Personalized
advertising to players
Distribution of game
assets in Unreal Engine
Throughput:Latency Spectrum
What's the
Secret?
TRIVIA
What do these have in common?
The Impact of Thread-Per-Core Architecture on Application Tail Latency
Fully Asynchronous Shard per Core
Shared-Nothing Architecture
Why ScyllaDB’s Shard Per Core Architecture Matters
Specialized Cache
■ Cache is LRU on rows
● Use BYPASS CACHE for analytical workloads
■ Unlimited Fetch Ceiling
■Efficient access & maintenance
●Thanks to replica collocation and design
CPU 0
CPU 1
CPU 2
CPU 3
Inside ScyllaDB's Internal Cache
API Compatibility: CQL and Amazon DynamoDB
Key Value Document Wide Column Graph
ScyllaDB vs. DynamoDB
We are ScyllaDB.
Not Cassandra, not DynamoDB.
Repair? Tombstones? Data Resurrection?
■ Worst things a database can do:
■ Lose data
■ Corrupt data
■ Resurrect data
■ Not a problem with ScyllaDB
■ We take your data seriously
■ We know repair is painful
Faster, Safer Node Operations with Repair vs Streaming
Tablets – True Elastic Scale
■ Faster Topology Changes
■ Immediate Request Serving
■ Easy Downscale (saves money :-)
■ Auto Balancing
AWS DynamoDB Auto Scaling is Not a Magic Bullet
Tablet File-based streaming
■ "Similar" to Cassandra Zero-copy Streaming
■ But actually works ;-)
■ Tablets are always owned by the replica
■ Simply copy, done.
■ 75% faster than Open Source for Streaming
Raft – Strongly Consistent *
■ Foundation for Strongly Consistent:
■ Tablets
■ Schema Changes
■ Topology Changes
■ Authentication Updates
■ Service Levels
RAFT
Safe
schema
changes
Safe
topology
changes
Dynamic
partitionin
g
Consistent
tables
Tablets
5.0
5.2
5.2+
6.0
ScyllaDB’s Safe Topology and Schema Changes on Raft
ScyllaDB Enterprise
■ Up to 1.5x Higher Throughput than Open Source
■ Up to 35% Lower Latencies (mean and P99)
New ScyllaDB Enterprise: Up to 50% Higher Throughput, 33% Lower Latency
Network (RPC) Compression Improvements
bit.ly/scylladb-getting-started
Get Started with ScyllaDB!
Next Steps
ScyllaDB University
Free online learning
scylladb.com/university
ScyllaDB
Resources
Visit our resource center
resources.scylladb.com
ScyllaDB Cloud
Start free trial
scylladb.com/cloud
Thank you
for joining us today.
@scylladb scylladb/
slack.scylladb.com
@scylladb company/scylladb/
scylladb/
cout << "Hello World!";
Felipe Mendes, Technical Director at ScyllaDB
+ Published Author on Linux and Databases
+ ScyllaDB Committer
+ "Plant a tree, have a child, write a book"
Poll: What are
your existing
challenges?
Poll: Which databases
are you currently
working with?
assertm(1 + 1 == 5, "Gotta run!");
Felipe Mendes
felipe@scylladb.com
@felipemendes.dev
scylladb.com
Thank you! Let’s connect.
We value your feedback!

ScyllaDB Virtual Workshop: Getting Started with ScyllaDB 2024

  • 1.
    Felipe Mendes, TechnicalDirector at ScyllaDB Getting Started with ScyllaDB
  • 2.
    Agenda + ModernScaling Challenges + What is ScyllaDB? + ScyllaDB Features + Getting Started & Next Steps
  • 3.
  • 4.
    Modern scaling challenges KeepCapEx & OpEx in check Reduce complexity Scale as the data grows Queries in milliseconds Leverage massive amounts of data Predictable, consistent performance
  • 5.
    Nestify – TheBusiness Case for Investing in Web Performance Optimization Time Matters
  • 6.
    Database Evolution OverTime Traditional RDBMS: + Structured data + Doesn’t scale Document Store: + No structure + Limited scale Key-Value Store: + Simple + Fast + Complex Column Store: + Some structure + Best HA/DR You probably use one or more of these for your organization :-)
  • 7.
    Lower Node Count Efficient hardwareusage results in direct savings What if we had it all in one… Predictable, Low Latencies Consistent single-digit millisecond p99 latencies Less Complexity Self-optimizing, smaller footprint, easy to use Developer-friendly APIs Ease of migration and less development cycles
  • 8.
  • 9.
  • 10.
  • 11.
    ScyllaDB and Memcached WeCompared ScyllaDB and Memcached and… We Lost?
  • 12.
    Trillions of messages 60%fewer nodes 5msp99 “It’s a much more efficient database — we’re going from running 177 Cassandra nodes to just 72 ScyllaDB nodes.” 2015 – Million 2016 – Billion 2019 – Trillion Bo Ingram How Discord Stores Billions and Trillions of Messages
  • 13.
    +400 Gamechangers LeverageScyllaDB Seamless experiences across content + devices Connecting people all around the globe Corporate fleet management Real-time analytics 2,000,000 SKU -commerce management Video recommendation management Elastic Kafka Persistent State Store at scale Real time fraud detection across 6M transactions/day Uber scale, mission critical chat & messaging app Network security threat detection Power ~50M X1 DVRs with billions of reqs/day Precision healthcare via Edison AI Inventory hub for retail operations Property listings and updates Unified ML feature store across the business Cryptocurrency exchange app Geography-based recommendations Global operations- Avon, Body Shop + more Predictable performance for on sale surges GPS-based exercise tracking Serving dynamic live streams at scale Powering India's top social media platform Personalized advertising to players Distribution of game assets in Unreal Engine
  • 14.
  • 15.
  • 16.
    TRIVIA What do thesehave in common?
  • 17.
    The Impact ofThread-Per-Core Architecture on Application Tail Latency
  • 18.
    Fully Asynchronous Shardper Core Shared-Nothing Architecture Why ScyllaDB’s Shard Per Core Architecture Matters
  • 19.
    Specialized Cache ■ Cacheis LRU on rows ● Use BYPASS CACHE for analytical workloads ■ Unlimited Fetch Ceiling ■Efficient access & maintenance ●Thanks to replica collocation and design CPU 0 CPU 1 CPU 2 CPU 3 Inside ScyllaDB's Internal Cache
  • 20.
    API Compatibility: CQLand Amazon DynamoDB Key Value Document Wide Column Graph ScyllaDB vs. DynamoDB
  • 21.
    We are ScyllaDB. NotCassandra, not DynamoDB.
  • 22.
    Repair? Tombstones? DataResurrection? ■ Worst things a database can do: ■ Lose data ■ Corrupt data ■ Resurrect data ■ Not a problem with ScyllaDB ■ We take your data seriously ■ We know repair is painful Faster, Safer Node Operations with Repair vs Streaming
  • 23.
    Tablets – TrueElastic Scale ■ Faster Topology Changes ■ Immediate Request Serving ■ Easy Downscale (saves money :-) ■ Auto Balancing AWS DynamoDB Auto Scaling is Not a Magic Bullet
  • 24.
    Tablet File-based streaming ■"Similar" to Cassandra Zero-copy Streaming ■ But actually works ;-) ■ Tablets are always owned by the replica ■ Simply copy, done. ■ 75% faster than Open Source for Streaming
  • 25.
    Raft – StronglyConsistent * ■ Foundation for Strongly Consistent: ■ Tablets ■ Schema Changes ■ Topology Changes ■ Authentication Updates ■ Service Levels RAFT Safe schema changes Safe topology changes Dynamic partitionin g Consistent tables Tablets 5.0 5.2 5.2+ 6.0 ScyllaDB’s Safe Topology and Schema Changes on Raft
  • 26.
    ScyllaDB Enterprise ■ Upto 1.5x Higher Throughput than Open Source ■ Up to 35% Lower Latencies (mean and P99) New ScyllaDB Enterprise: Up to 50% Higher Throughput, 33% Lower Latency
  • 27.
  • 28.
  • 29.
    Next Steps ScyllaDB University Freeonline learning scylladb.com/university ScyllaDB Resources Visit our resource center resources.scylladb.com ScyllaDB Cloud Start free trial scylladb.com/cloud
  • 30.
    Thank you for joiningus today. @scylladb scylladb/ slack.scylladb.com @scylladb company/scylladb/ scylladb/
  • 31.
    cout << "HelloWorld!"; Felipe Mendes, Technical Director at ScyllaDB + Published Author on Linux and Databases + ScyllaDB Committer + "Plant a tree, have a child, write a book"
  • 32.
    Poll: What are yourexisting challenges?
  • 33.
    Poll: Which databases areyou currently working with?
  • 34.
    assertm(1 + 1== 5, "Gotta run!"); Felipe Mendes felipe@scylladb.com @felipemendes.dev scylladb.com Thank you! Let’s connect.
  • 35.
    We value yourfeedback!