SlideShare a Scribd company logo
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Scylla on Samsung Z-SSD
Senior Performance Architect, Samsung Semiconductor Inc.
Arash Rezaei
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Arash Rezaei
2
Arash is a senior performance architect at Samsung
Semiconductor Inc. His work has been primarily on
performance evaluation and tuning of data center
applications with a focus on storage.
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Agenda
3
▪ Z-SSD – Samsung’s ultra-low latency SSD
▪ Z-SSD use-cases
▪ Scylla on Z-SSD
▪ Summary
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Introduction: Latency Matters
4
▪ Threshold of patience is decreasing
o Retrieving data from caches, databases, data stores
o Business analytics
MySQL/ NoSQL
Databases
Streaming
Data
Storage
Layer
Cache Layer
Web
Layer
Analytics
Latency

Recommended for you

Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot InstancesScylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances

Scylla and Spotinst together provide a strong combination of extreme performance and cost reduction. In this talk, we will present how a Scylla cluster can be used on AWS’s EC2 Spot without losing consistency with the help of Spotinst prediction technology and advanced stateful features. We will show a live demo on how to run Scylla on the Spotinst platform.

nosqlscyllasummitscylla
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...

This document outlines a presentation on using the GoCQL driver to execute queries against Cassandra and Scylla databases. It discusses connecting to a Cassandra cluster, executing queries, iterating over results, and using asynchronous queries. It also mentions some additional Cassandra libraries built on top of GoCQL, including gocqlx for data binding and queries, and gocassa for queries and migrations. The presentation aims to explain how GoCQL works behind the scenes and how to get started with basic querying functionality.

scyllanosqlscyllasummit
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...

Benchmarks are fun to do but when going to production, all sorts of things can happen: anything from hardware outages to human error bringing your database down. Even in a healthy database, a lot of maintenance operations have to periodically run. Do you have the tools necessary to make sure you are good to go?

nosqlscyllasummitscylla
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Samsung Z-SSD
5
▪ Ultra-low latency flash storage device
▪ Provides 5.5x lower latency comparing to today’s leading NVMe SSD
14.7
80.3
0 10 20 30 40 50 60 70 80 90
Z-SSD
PM1725a
Latency [usec]
4K Random
Read Latency
5.5x Lower Latency
fio --rw=randread --bs=4k --direct=1 --iodepth=1 –numjobs=1 --
ioengine=libaio --norandommap
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Samsung Z-SSD
6
DRAM
Z-NAND
New Controller
50%▼
Past LP4
70%▼
New CTRLPast
x 1/10 ▼
Z-NANDPast
▪ Samsung’s Z-SSD is vertically optimized from DRAM to storage
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
SZ985: Samsung Z-SSD
7
▪ Sequential traffic
o 3.2 GB/sec Read
o 3.2 GB/sec Write
▪ Random traffic
o 750K IOPs Reads
o 170K IOPs Writes
▪ Capacity
o 800GB
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Z-SSD Minimizes Latency
8
▪ High performance @ consistent low latency
SSD
Throughput Latency
Sequential
R/W
Random
R/W
Random
Read
PM1725a
3.3/2.95
GB/s
800/160
KIOPS
90µs
Z-SSD 3.2/3.2 GB/s
750/170
KIOPs
15µs
Read-only workload (FIO) Mixed (80%/20% R/W) FIO

Recommended for you

Scylla Summit 2017: Planning Your Queries for Maximum Performance
Scylla Summit 2017: Planning Your Queries for Maximum PerformanceScylla Summit 2017: Planning Your Queries for Maximum Performance
Scylla Summit 2017: Planning Your Queries for Maximum Performance

What happens to a request that reaches Scylla, and why should one care? Understanding how Scylla executes your queries can help you make better architectural decisions and also better understand the performance of your application. Are my rows too big? Should I make that other column a part of my partition key instead? This talk will cover the interaction between nodes, shards and the role of Scylla's internal components like memtables, cache and sstables. I will explain how different types of queries are executed and how to plan your queries for maximum performance.

scyllasummitnosqlscylla
Scylla Summit 2017: A Toolbox for Understanding Scylla in the Field
Scylla Summit 2017: A Toolbox for Understanding Scylla in the FieldScylla Summit 2017: A Toolbox for Understanding Scylla in the Field
Scylla Summit 2017: A Toolbox for Understanding Scylla in the Field

In this talk, we will share useful tools and techniques that we are using in the field to understand Scylla clusters. Users will learn how to use those same tools to better understand their deployment. Some of the questions that will be answered are: - how to find out which queries are the slowest and why - how we go about understanding the impact of the data model in a node's performance - how to check which resources are the bottlenecks in the cluster

nosqlscyllasummitscylla
If You Care About Performance, Use User Defined Types
If You Care About Performance, Use User Defined TypesIf You Care About Performance, Use User Defined Types
If You Care About Performance, Use User Defined Types

Shlomi Livne, VP of R&D at ScyllaDB, presented on the performance benefits of using user-defined types (UDTs) in ScyllaDB. He explained that with traditional columns, each column has overhead and flexibility comes at a price. However, with frozen UDTs, the columns are treated as a single unit, sharing metadata and improving performance. Livne showed results of a test where UDTs with many fields outperformed traditional columns with the same number of fields. However, he noted that Scylla's row cache and Java driver performance need improvement for UDTs.

nosqlscyllasummitscylla
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Z-SSD Use Cases
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Z-SSD Benefits Latency-Sensitive Apps (1/2)
10
▪ Z-SSD as swap space
▪ Memcached: distributed
object caching system
▪ 3x speedup @ 1
3 the
average latency under
heavy swap load
0
20000
40000
60000
80000
100000
memcached
Throughput[op/sec]
Memcached
Throughput
PM1725a Z-SSD
0
200
400
600
800
1000
1200
1400
memcached
ReadLatency[usec]
Memcached
Latency
PM1725a Z-SSD
3x
𝟏
𝟑
memcached with 60GB dataset (45 million records, value size=1000B)
DRAM capacity=32GB (less than half of the used memory space)
Driven with YCSB, 90% read/ 10% update
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Z-SSD Benefits Latency-Sensitive Apps (2/2)
11
▪ Z-SSD as storage layer
▪ RocksDB is used as storage
engine for DBs
▪ 2x throughput @ ½ the
average latency
0
20000
40000
60000
80000
100000
120000
140000
RocksDB
Throughput[op/sec]
RocksDB
Throughput
PM1725a Z-SSD
2x
0
50
100
150
200
250
300
350
RocksDB
AverateLatency[usec]
RocksDB
Average Latency
PM1725a Z-SSD
½
db_bench read-while-writing workload
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Scylla on Z-SSD

Recommended for you

Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark

When working with streaming data, stateful operations are a common use case. If you would like to perform data de-duplication, calculate aggregations over event-time windows, track user activity over sessions, you are performing a stateful operation. Apache Spark provides users with a high level, simple to use DataFrame/Dataset API to work with both batch and streaming data. The funny thing about batch workloads is that people tend to run these batch workloads over and over again. Structured Streaming allows users to run these same workloads, with the exact same business logic in a streaming fashion, helping users answer questions at lower latencies. In this talk, we will focus on stateful operations with Structured Streaming and we will demonstrate through live demos, how NoSQL stores can be plugged in as a fault tolerant state store to store intermediate state, as well as used as a streaming sink, where the output data can be stored indefinitely for downstream applications.

nosqlscyllasummitscylla
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of ViewScylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View

Are you a MySQL DBA or DevOps individual being asked to run Cassandra or Scylla? Feeling overwhelmed? In this talk, I will present Cassandra/Scylla operations in terms that directly relate to MySQL. I will show you comparisons between the Information Schema and the Cassandra/Scylla System keyspace(s). I will also talk about metrics available in MySQL versus Cassandra/Scylla and how to retrieve them. Finally, I will talk about how MySQL replication compares with Cassandra replication. Hopefully, when I am done you will be able to relate to Cassandra operations in a practical and useful way.

nosqlscyllasummitscylladb
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...

We will share Scylla adoption practices in equipment sensor data management of MES, Data Modeling Tips, Data Architecture using Scylla, configurations, and tunings.

scyllascyllasummitnosql
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Methodology
13
▪ Goal:
o Compare Scylla in-memory vs. Scylla on Z-SSD
▪ Metric:
o Max throughput under given tail-latency constraints
▪ Workload:
o Cassandra-stress with:
• Varying Read/Write ratio (100%/0% – 75/25% )
• Varying % of request served from memory (25% to 75%)
o Database populated with 300GB data (payload 1KB)
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
System Configuration
14
Item Description
CPU 2 x Intel Xeon Gold 6154 @
3.00GHz (total 36 cores per node)
RAM 128GB
SSD 2 x SZ985 800GB
OS Ubuntu 16.04.2, Kernel 4.12.0
Scylla version 2.0
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Throughput while Meeting Latency SLA (1/2)
▪ 100% read workload
15
Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Throughput while Meeting Latency SLA (1/2)
▪ 100% read workload
16
Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency

Recommended for you

Scylla Summit 2017: Keynote, Looking back, looking ahead
Scylla Summit 2017: Keynote, Looking back, looking aheadScylla Summit 2017: Keynote, Looking back, looking ahead
Scylla Summit 2017: Keynote, Looking back, looking ahead

ScyllaDB CTO Avi Kivity gave a keynote on how Scylla has evolved. He discussed new features in Scylla 2.0—including Materialized Views and Heat-Weighted Load Balancing, changes in monitoring—and shared our product roadmap. He also talked about our recent acquisition of Seastar.io and how it will enable us to deliver a database-as-a-service offering.

scyllanosqlscyllasummit
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPS
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPSScylla Summit 2017: Running a Soft Real-time Service at One Million QPS
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPS

AdGear runs an ad tech gateway at more than one million queries per second to Scylla and recently transitioned from Apache Cassandra. In this talk, we will highlight the tools and languages that we use (Erlang), how we do bulk imports, and how performance compares between the two database engines.

scylladbscyllasummitnosql
Scylla Summit 2017: Scylla on Kubernetes
Scylla Summit 2017: Scylla on KubernetesScylla Summit 2017: Scylla on Kubernetes
Scylla Summit 2017: Scylla on Kubernetes

Kubernetes is a declarative system for automatically deploying, managing, and scaling applications and their dependencies. In this short talk, I'll demonstrate a small Scylla cluster running in Google Compute Engine via Kubernetes and our publicly-published Docker images.

scyllasummitnosqlscylla
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Throughput while Meeting Latency SLA (2/2)
▪ 75% read/25% write workload
17
Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Throughput while Meeting Latency SLA (2/2)
▪ 75% read/25% write workload
18
Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
19
▪ Server viewpoint
▪ Latency under a fixed throughput
▪ 100% read workload
(95 percentile latency)
Software Overheads
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
20
▪ Server viewpoint
▪ Latency under a fixed throughput
▪ 100% read workload
(95 percentile latency)
Software Overheads
Device tail latency from FIO experiments

Recommended for you

Scylla Summit 2017: The Upcoming HPC Evolution
Scylla Summit 2017: The Upcoming HPC EvolutionScylla Summit 2017: The Upcoming HPC Evolution
Scylla Summit 2017: The Upcoming HPC Evolution

In this talk, I will explain how HPC is beginning to evolve and how we use supercomputers to monitor supercomputers. First we will look at how HPC is different from cloud computing in terms of infrastructure and application architecture. Then I will discuss how those things are changing and why. Finally, I will dive into a use case of monitoring supercomputers as an application area for Scylla.

scyllasummitscyllanosql
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...

The document appears to be a presentation on optimizing inter-data center communication. It discusses key topics like what inter-data center communication involves, the costs associated with it, best practices for setting snitches, keyspaces, client drivers and consistency levels for queries to optimize performance between data centers. It recommends using network topology replication strategies over simple strategies for multi-region deployments, setting load balancing and consistency levels appropriately in clients, and enabling internode compression to reduce costs of communication between data centers. The presentation encourages reviewing client locations, data access patterns, who is reading/writing data, and having conversations between operations and development teams to determine the best use cases.

nosqlscyllasummitscylla
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...

JanusGraph, a highly scalable graph database solution, supports historically Cassandra and HBase as database backends. We decided to put Scylla in the mix, certainly searching for the best performing backend. We ran test scenarios that cover high volume reads and writes. In this talk, we will show you the performance results of Scylla vs others and also share our lessons learned during the performance evaluation.

scylladbscyllasummitnosql
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Scylla: DRAM, Z-SSD, or NVMe SSD?
21
▪ For a given use-case requirement (throughput and DB size), what is
the most cost-efficient cluster configuration?
▪ Optimize performance/$ for different use-cases
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
Summary
▪ Samsung Z-SSD: ultra-low latency SSD device
o 5.5x lower latency compared to NVMe SSD
▪ Z-SSD benefits latency-sensitive applications
o RocksDB: 2x throughput @ ½ the average latency
o Memcached (as swap): 3x speedup @ 1
3 the average latency
▪ Z-SSD for Scylla:
o Helps reduce the performance gap between memory and storage
• Delivers 44% – 23% of DRAM performance at significant lower cost
– 50% – 75% of requests served from memory
o Software improvements could further shrink the memory/storage gap
o Opens new options for architecting a large cluster
22
PRESENTATION TITLE ON ONE LINE
AND ON TWO LINES
First and last name
Position, company
THANK YOU
a.rezaei@samsung.com
Please stay in touch
Any questions?

More Related Content

What's hot

Scylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data Center
Scylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data CenterScylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data Center
Scylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data Center
ScyllaDB
 
Scylla Summit 2017: SMF: The Fastest RPC in the West
Scylla Summit 2017: SMF: The Fastest RPC in the WestScylla Summit 2017: SMF: The Fastest RPC in the West
Scylla Summit 2017: SMF: The Fastest RPC in the West
ScyllaDB
 
Scylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor Laor
Scylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor LaorScylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor Laor
Scylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor Laor
ScyllaDB
 
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot InstancesScylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
ScyllaDB
 
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
ScyllaDB
 
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
ScyllaDB
 
Scylla Summit 2017: Planning Your Queries for Maximum Performance
Scylla Summit 2017: Planning Your Queries for Maximum PerformanceScylla Summit 2017: Planning Your Queries for Maximum Performance
Scylla Summit 2017: Planning Your Queries for Maximum Performance
ScyllaDB
 
Scylla Summit 2017: A Toolbox for Understanding Scylla in the Field
Scylla Summit 2017: A Toolbox for Understanding Scylla in the FieldScylla Summit 2017: A Toolbox for Understanding Scylla in the Field
Scylla Summit 2017: A Toolbox for Understanding Scylla in the Field
ScyllaDB
 
If You Care About Performance, Use User Defined Types
If You Care About Performance, Use User Defined TypesIf You Care About Performance, Use User Defined Types
If You Care About Performance, Use User Defined Types
ScyllaDB
 
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
ScyllaDB
 
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of ViewScylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
ScyllaDB
 
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
ScyllaDB
 
Scylla Summit 2017: Keynote, Looking back, looking ahead
Scylla Summit 2017: Keynote, Looking back, looking aheadScylla Summit 2017: Keynote, Looking back, looking ahead
Scylla Summit 2017: Keynote, Looking back, looking ahead
ScyllaDB
 
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPS
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPSScylla Summit 2017: Running a Soft Real-time Service at One Million QPS
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPS
ScyllaDB
 
Scylla Summit 2017: Scylla on Kubernetes
Scylla Summit 2017: Scylla on KubernetesScylla Summit 2017: Scylla on Kubernetes
Scylla Summit 2017: Scylla on Kubernetes
ScyllaDB
 
Scylla Summit 2017: The Upcoming HPC Evolution
Scylla Summit 2017: The Upcoming HPC EvolutionScylla Summit 2017: The Upcoming HPC Evolution
Scylla Summit 2017: The Upcoming HPC Evolution
ScyllaDB
 
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
ScyllaDB
 
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
ScyllaDB
 
Scylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny Schnaider
Scylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny SchnaiderScylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny Schnaider
Scylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny Schnaider
ScyllaDB
 
Scylla Summit 2017: Welcome and Keynote - Nextgen NoSQL
Scylla Summit 2017: Welcome and Keynote - Nextgen NoSQLScylla Summit 2017: Welcome and Keynote - Nextgen NoSQL
Scylla Summit 2017: Welcome and Keynote - Nextgen NoSQL
ScyllaDB
 

What's hot (20)

Scylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data Center
Scylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data CenterScylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data Center
Scylla Summit 2017: Intel Optane SSDs as the New Accelerator in Your Data Center
 
Scylla Summit 2017: SMF: The Fastest RPC in the West
Scylla Summit 2017: SMF: The Fastest RPC in the WestScylla Summit 2017: SMF: The Fastest RPC in the West
Scylla Summit 2017: SMF: The Fastest RPC in the West
 
Scylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor Laor
Scylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor LaorScylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor Laor
Scylla Summit 2017 Keynote: NextGen NoSQL with CEO Dor Laor
 
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot InstancesScylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
Scylla Summit 2017: Saving Thousands by Running Scylla on EC2 Spot Instances
 
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
Scylla Summit 2017: How to Use Gocql to Execute Queries and What the Driver D...
 
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
Scylla Summit 2017: Repair, Backup, Restore: Last Thing Before You Go to Prod...
 
Scylla Summit 2017: Planning Your Queries for Maximum Performance
Scylla Summit 2017: Planning Your Queries for Maximum PerformanceScylla Summit 2017: Planning Your Queries for Maximum Performance
Scylla Summit 2017: Planning Your Queries for Maximum Performance
 
Scylla Summit 2017: A Toolbox for Understanding Scylla in the Field
Scylla Summit 2017: A Toolbox for Understanding Scylla in the FieldScylla Summit 2017: A Toolbox for Understanding Scylla in the Field
Scylla Summit 2017: A Toolbox for Understanding Scylla in the Field
 
If You Care About Performance, Use User Defined Types
If You Care About Performance, Use User Defined TypesIf You Care About Performance, Use User Defined Types
If You Care About Performance, Use User Defined Types
 
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
Scylla Summit 2017: Stateful Streaming Applications with Apache Spark
 
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of ViewScylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
Scylla Summit 2017: How to Run Cassandra/Scylla from a MySQL DBA's Point of View
 
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
Scylla Summit 2017: Scylla for Mass Simultaneous Sensor Data Processing of ME...
 
Scylla Summit 2017: Keynote, Looking back, looking ahead
Scylla Summit 2017: Keynote, Looking back, looking aheadScylla Summit 2017: Keynote, Looking back, looking ahead
Scylla Summit 2017: Keynote, Looking back, looking ahead
 
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPS
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPSScylla Summit 2017: Running a Soft Real-time Service at One Million QPS
Scylla Summit 2017: Running a Soft Real-time Service at One Million QPS
 
Scylla Summit 2017: Scylla on Kubernetes
Scylla Summit 2017: Scylla on KubernetesScylla Summit 2017: Scylla on Kubernetes
Scylla Summit 2017: Scylla on Kubernetes
 
Scylla Summit 2017: The Upcoming HPC Evolution
Scylla Summit 2017: The Upcoming HPC EvolutionScylla Summit 2017: The Upcoming HPC Evolution
Scylla Summit 2017: The Upcoming HPC Evolution
 
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
Scylla Summit 2017: How to Optimize and Reduce Inter-DC Network Traffic and S...
 
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
Scylla Summit 2017: Performance Evaluation of Scylla as a Database Backend fo...
 
Scylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny Schnaider
Scylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny SchnaiderScylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny Schnaider
Scylla Summit 2017 Keynote: NextGen NoSQL with Chairman Benny Schnaider
 
Scylla Summit 2017: Welcome and Keynote - Nextgen NoSQL
Scylla Summit 2017: Welcome and Keynote - Nextgen NoSQLScylla Summit 2017: Welcome and Keynote - Nextgen NoSQL
Scylla Summit 2017: Welcome and Keynote - Nextgen NoSQL
 

Viewers also liked

Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...
Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...
Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...
ScyllaDB
 
Scylla Summit 2017: Managing 10,000 Node Storage Clusters at Twitter
Scylla Summit 2017: Managing 10,000 Node Storage Clusters at TwitterScylla Summit 2017: Managing 10,000 Node Storage Clusters at Twitter
Scylla Summit 2017: Managing 10,000 Node Storage Clusters at Twitter
ScyllaDB
 
Scylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQL
Scylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQLScylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQL
Scylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQL
ScyllaDB
 
How to Monitor and Size Workloads on AWS i3 instances
How to Monitor and Size Workloads on AWS i3 instancesHow to Monitor and Size Workloads on AWS i3 instances
How to Monitor and Size Workloads on AWS i3 instances
ScyllaDB
 
Scylla Summit 2016: Keynote - Big Data Goes Native
Scylla Summit 2016: Keynote - Big Data Goes NativeScylla Summit 2016: Keynote - Big Data Goes Native
Scylla Summit 2016: Keynote - Big Data Goes Native
ScyllaDB
 
How to achieve no compromise performance and availability
How to achieve no compromise performance and availabilityHow to achieve no compromise performance and availability
How to achieve no compromise performance and availability
ScyllaDB
 

Viewers also liked (6)

Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...
Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...
Scylla Summit 2017: Cry in the Dojo, Laugh in the Battlefield: How We Constan...
 
Scylla Summit 2017: Managing 10,000 Node Storage Clusters at Twitter
Scylla Summit 2017: Managing 10,000 Node Storage Clusters at TwitterScylla Summit 2017: Managing 10,000 Node Storage Clusters at Twitter
Scylla Summit 2017: Managing 10,000 Node Storage Clusters at Twitter
 
Scylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQL
Scylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQLScylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQL
Scylla Summit 2017: Streaming ETL in Kafka for Everyone with KSQL
 
How to Monitor and Size Workloads on AWS i3 instances
How to Monitor and Size Workloads on AWS i3 instancesHow to Monitor and Size Workloads on AWS i3 instances
How to Monitor and Size Workloads on AWS i3 instances
 
Scylla Summit 2016: Keynote - Big Data Goes Native
Scylla Summit 2016: Keynote - Big Data Goes NativeScylla Summit 2016: Keynote - Big Data Goes Native
Scylla Summit 2016: Keynote - Big Data Goes Native
 
How to achieve no compromise performance and availability
How to achieve no compromise performance and availabilityHow to achieve no compromise performance and availability
How to achieve no compromise performance and availability
 

Similar to Scylla Summit 2017: Scylla on Samsung NVMe Z-SSDs

오라클 DR 및 복제 솔루션(Dbvisit 소개)
오라클 DR 및 복제 솔루션(Dbvisit 소개)오라클 DR 및 복제 솔루션(Dbvisit 소개)
오라클 DR 및 복제 솔루션(Dbvisit 소개)
Linux Foundation Korea
 
Orcl siebel-sun-s282213-oow2006
Orcl siebel-sun-s282213-oow2006Orcl siebel-sun-s282213-oow2006
Orcl siebel-sun-s282213-oow2006
Sal Marcus
 
Oracle RAC 12c Overview
Oracle RAC 12c OverviewOracle RAC 12c Overview
Oracle RAC 12c Overview
Markus Michalewicz
 
Collaborate07kmohiuddin
Collaborate07kmohiuddinCollaborate07kmohiuddin
Collaborate07kmohiuddin
Sal Marcus
 
Understanding Oracle RAC 12c Internals OOW13 [CON8806]
Understanding Oracle RAC 12c Internals OOW13 [CON8806]Understanding Oracle RAC 12c Internals OOW13 [CON8806]
Understanding Oracle RAC 12c Internals OOW13 [CON8806]
Markus Michalewicz
 
gDBClone - Database Clone “onecommand Automation Tool”
gDBClone - Database Clone “onecommand Automation Tool”gDBClone - Database Clone “onecommand Automation Tool”
gDBClone - Database Clone “onecommand Automation Tool”
Ruggero Citton
 
DAT316_Report from the field on Aurora PostgreSQL Performance
DAT316_Report from the field on Aurora PostgreSQL PerformanceDAT316_Report from the field on Aurora PostgreSQL Performance
DAT316_Report from the field on Aurora PostgreSQL Performance
Amazon Web Services
 
Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...
Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...
Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...
Amazon Web Services
 
Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28
Amazon Web Services
 
PostgreSQL replication
PostgreSQL replicationPostgreSQL replication
PostgreSQL replication
NTT DATA OSS Professional Services
 
What's new in Amazon Aurora - ADB203 - Chicago AWS Summit
What's new in Amazon Aurora - ADB203 - Chicago AWS SummitWhat's new in Amazon Aurora - ADB203 - Chicago AWS Summit
What's new in Amazon Aurora - ADB203 - Chicago AWS Summit
Amazon Web Services
 
Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집
Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집
Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집
AWSKRUG - AWS한국사용자모임
 
DAT202_Getting started with Amazon Aurora
DAT202_Getting started with Amazon AuroraDAT202_Getting started with Amazon Aurora
DAT202_Getting started with Amazon Aurora
Amazon Web Services
 
5 Things You Need to Know About Enterprise Fl
 5 Things You Need to Know About Enterprise Fl 5 Things You Need to Know About Enterprise Fl
5 Things You Need to Know About Enterprise Fl
Western Digital
 
AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)
AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)
AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)
Amazon Web Services
 
How to Upgrade Major Version of Your Production PostgreSQL
How to Upgrade Major Version of Your Production PostgreSQLHow to Upgrade Major Version of Your Production PostgreSQL
How to Upgrade Major Version of Your Production PostgreSQL
Keisuke Suzuki
 
Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019
Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019
Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019
javier ramirez
 
Tech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of Facebook
Tech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of FacebookTech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of Facebook
Tech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of Facebook
The Hive
 
EMCSymmetrix vmax-10
EMCSymmetrix vmax-10EMCSymmetrix vmax-10
EMCSymmetrix vmax-10
solarisyougood
 
Presentation implementing oracle asm successfully
Presentation    implementing oracle asm successfullyPresentation    implementing oracle asm successfully
Presentation implementing oracle asm successfully
xKinAnx
 

Similar to Scylla Summit 2017: Scylla on Samsung NVMe Z-SSDs (20)

오라클 DR 및 복제 솔루션(Dbvisit 소개)
오라클 DR 및 복제 솔루션(Dbvisit 소개)오라클 DR 및 복제 솔루션(Dbvisit 소개)
오라클 DR 및 복제 솔루션(Dbvisit 소개)
 
Orcl siebel-sun-s282213-oow2006
Orcl siebel-sun-s282213-oow2006Orcl siebel-sun-s282213-oow2006
Orcl siebel-sun-s282213-oow2006
 
Oracle RAC 12c Overview
Oracle RAC 12c OverviewOracle RAC 12c Overview
Oracle RAC 12c Overview
 
Collaborate07kmohiuddin
Collaborate07kmohiuddinCollaborate07kmohiuddin
Collaborate07kmohiuddin
 
Understanding Oracle RAC 12c Internals OOW13 [CON8806]
Understanding Oracle RAC 12c Internals OOW13 [CON8806]Understanding Oracle RAC 12c Internals OOW13 [CON8806]
Understanding Oracle RAC 12c Internals OOW13 [CON8806]
 
gDBClone - Database Clone “onecommand Automation Tool”
gDBClone - Database Clone “onecommand Automation Tool”gDBClone - Database Clone “onecommand Automation Tool”
gDBClone - Database Clone “onecommand Automation Tool”
 
DAT316_Report from the field on Aurora PostgreSQL Performance
DAT316_Report from the field on Aurora PostgreSQL PerformanceDAT316_Report from the field on Aurora PostgreSQL Performance
DAT316_Report from the field on Aurora PostgreSQL Performance
 
Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...
Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...
Report from the Field on the PostgreSQL-compatible Edition of Amazon Aurora -...
 
Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28
 
PostgreSQL replication
PostgreSQL replicationPostgreSQL replication
PostgreSQL replication
 
What's new in Amazon Aurora - ADB203 - Chicago AWS Summit
What's new in Amazon Aurora - ADB203 - Chicago AWS SummitWhat's new in Amazon Aurora - ADB203 - Chicago AWS Summit
What's new in Amazon Aurora - ADB203 - Chicago AWS Summit
 
Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집
Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집
Aurora Serverless, 서버리스 RDB의 서막 - 트랙2, Community Day 2018 re:Invent 특집
 
DAT202_Getting started with Amazon Aurora
DAT202_Getting started with Amazon AuroraDAT202_Getting started with Amazon Aurora
DAT202_Getting started with Amazon Aurora
 
5 Things You Need to Know About Enterprise Fl
 5 Things You Need to Know About Enterprise Fl 5 Things You Need to Know About Enterprise Fl
5 Things You Need to Know About Enterprise Fl
 
AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)
AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)
AWS re:Invent 2016: Getting Started with Amazon Aurora (DAT203)
 
How to Upgrade Major Version of Your Production PostgreSQL
How to Upgrade Major Version of Your Production PostgreSQLHow to Upgrade Major Version of Your Production PostgreSQL
How to Upgrade Major Version of Your Production PostgreSQL
 
Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019
Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019
Scalable Relational Databases with Amazon Aurora. Madrid Summit 2019
 
Tech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of Facebook
Tech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of FacebookTech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of Facebook
Tech Talk: RocksDB Slides by Dhruba Borthakur & Haobo Xu of Facebook
 
EMCSymmetrix vmax-10
EMCSymmetrix vmax-10EMCSymmetrix vmax-10
EMCSymmetrix vmax-10
 
Presentation implementing oracle asm successfully
Presentation    implementing oracle asm successfullyPresentation    implementing oracle asm successfully
Presentation implementing oracle asm successfully
 

More from ScyllaDB

Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...
Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...
Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...
ScyllaDB
 
Mitigating the Impact of State Management in Cloud Stream Processing Systems
Mitigating the Impact of State Management in Cloud Stream Processing SystemsMitigating the Impact of State Management in Cloud Stream Processing Systems
Mitigating the Impact of State Management in Cloud Stream Processing Systems
ScyllaDB
 
Measuring the Impact of Network Latency at Twitter
Measuring the Impact of Network Latency at TwitterMeasuring the Impact of Network Latency at Twitter
Measuring the Impact of Network Latency at Twitter
ScyllaDB
 
Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...
Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...
Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...
ScyllaDB
 
Noise Canceling RUM by Tim Vereecke, Akamai
Noise Canceling RUM by Tim Vereecke, AkamaiNoise Canceling RUM by Tim Vereecke, Akamai
Noise Canceling RUM by Tim Vereecke, Akamai
ScyllaDB
 
Running a Go App in Kubernetes: CPU Impacts
Running a Go App in Kubernetes: CPU ImpactsRunning a Go App in Kubernetes: CPU Impacts
Running a Go App in Kubernetes: CPU Impacts
ScyllaDB
 
Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...
Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...
Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...
ScyllaDB
 
Performance Budgets for the Real World by Tammy Everts
Performance Budgets for the Real World by Tammy EvertsPerformance Budgets for the Real World by Tammy Everts
Performance Budgets for the Real World by Tammy Everts
ScyllaDB
 
Using Libtracecmd to Analyze Your Latency and Performance Troubles
Using Libtracecmd to Analyze Your Latency and Performance TroublesUsing Libtracecmd to Analyze Your Latency and Performance Troubles
Using Libtracecmd to Analyze Your Latency and Performance Troubles
ScyllaDB
 
Reducing P99 Latencies with Generational ZGC
Reducing P99 Latencies with Generational ZGCReducing P99 Latencies with Generational ZGC
Reducing P99 Latencies with Generational ZGC
ScyllaDB
 
5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X
5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X
5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X
ScyllaDB
 
How Netflix Builds High Performance Applications at Global Scale
How Netflix Builds High Performance Applications at Global ScaleHow Netflix Builds High Performance Applications at Global Scale
How Netflix Builds High Performance Applications at Global Scale
ScyllaDB
 
Conquering Load Balancing: Experiences from ScyllaDB Drivers
Conquering Load Balancing: Experiences from ScyllaDB DriversConquering Load Balancing: Experiences from ScyllaDB Drivers
Conquering Load Balancing: Experiences from ScyllaDB Drivers
ScyllaDB
 
Interaction Latency: Square's User-Centric Mobile Performance Metric
Interaction Latency: Square's User-Centric Mobile Performance MetricInteraction Latency: Square's User-Centric Mobile Performance Metric
Interaction Latency: Square's User-Centric Mobile Performance Metric
ScyllaDB
 
How to Avoid Learning the Linux-Kernel Memory Model
How to Avoid Learning the Linux-Kernel Memory ModelHow to Avoid Learning the Linux-Kernel Memory Model
How to Avoid Learning the Linux-Kernel Memory Model
ScyllaDB
 
99.99% of Your Traces are Trash by Paige Cruz
99.99% of Your Traces are Trash by Paige Cruz99.99% of Your Traces are Trash by Paige Cruz
99.99% of Your Traces are Trash by Paige Cruz
ScyllaDB
 
Square's Lessons Learned from Implementing a Key-Value Store with Raft
Square's Lessons Learned from Implementing a Key-Value Store with RaftSquare's Lessons Learned from Implementing a Key-Value Store with Raft
Square's Lessons Learned from Implementing a Key-Value Store with Raft
ScyllaDB
 
Making Python 100x Faster with Less Than 100 Lines of Rust
Making Python 100x Faster with Less Than 100 Lines of RustMaking Python 100x Faster with Less Than 100 Lines of Rust
Making Python 100x Faster with Less Than 100 Lines of Rust
ScyllaDB
 
A Deep Dive Into Concurrent React by Matheus Albuquerque
A Deep Dive Into Concurrent React by Matheus AlbuquerqueA Deep Dive Into Concurrent React by Matheus Albuquerque
A Deep Dive Into Concurrent React by Matheus Albuquerque
ScyllaDB
 
The Latency Stack: Discovering Surprising Sources of Latency
The Latency Stack: Discovering Surprising Sources of LatencyThe Latency Stack: Discovering Surprising Sources of Latency
The Latency Stack: Discovering Surprising Sources of Latency
ScyllaDB
 

More from ScyllaDB (20)

Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...
Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...
Unconventional Methods to Identify Bottlenecks in Low-Latency and High-Throug...
 
Mitigating the Impact of State Management in Cloud Stream Processing Systems
Mitigating the Impact of State Management in Cloud Stream Processing SystemsMitigating the Impact of State Management in Cloud Stream Processing Systems
Mitigating the Impact of State Management in Cloud Stream Processing Systems
 
Measuring the Impact of Network Latency at Twitter
Measuring the Impact of Network Latency at TwitterMeasuring the Impact of Network Latency at Twitter
Measuring the Impact of Network Latency at Twitter
 
Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...
Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...
Architecting a High-Performance (Open Source) Distributed Message Queuing Sys...
 
Noise Canceling RUM by Tim Vereecke, Akamai
Noise Canceling RUM by Tim Vereecke, AkamaiNoise Canceling RUM by Tim Vereecke, Akamai
Noise Canceling RUM by Tim Vereecke, Akamai
 
Running a Go App in Kubernetes: CPU Impacts
Running a Go App in Kubernetes: CPU ImpactsRunning a Go App in Kubernetes: CPU Impacts
Running a Go App in Kubernetes: CPU Impacts
 
Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...
Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...
Always-on Profiling of All Linux Threads, On-CPU and Off-CPU, with eBPF & Con...
 
Performance Budgets for the Real World by Tammy Everts
Performance Budgets for the Real World by Tammy EvertsPerformance Budgets for the Real World by Tammy Everts
Performance Budgets for the Real World by Tammy Everts
 
Using Libtracecmd to Analyze Your Latency and Performance Troubles
Using Libtracecmd to Analyze Your Latency and Performance TroublesUsing Libtracecmd to Analyze Your Latency and Performance Troubles
Using Libtracecmd to Analyze Your Latency and Performance Troubles
 
Reducing P99 Latencies with Generational ZGC
Reducing P99 Latencies with Generational ZGCReducing P99 Latencies with Generational ZGC
Reducing P99 Latencies with Generational ZGC
 
5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X
5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X
5 Hours to 7.7 Seconds: How Database Tricks Sped up Rust Linting Over 2000X
 
How Netflix Builds High Performance Applications at Global Scale
How Netflix Builds High Performance Applications at Global ScaleHow Netflix Builds High Performance Applications at Global Scale
How Netflix Builds High Performance Applications at Global Scale
 
Conquering Load Balancing: Experiences from ScyllaDB Drivers
Conquering Load Balancing: Experiences from ScyllaDB DriversConquering Load Balancing: Experiences from ScyllaDB Drivers
Conquering Load Balancing: Experiences from ScyllaDB Drivers
 
Interaction Latency: Square's User-Centric Mobile Performance Metric
Interaction Latency: Square's User-Centric Mobile Performance MetricInteraction Latency: Square's User-Centric Mobile Performance Metric
Interaction Latency: Square's User-Centric Mobile Performance Metric
 
How to Avoid Learning the Linux-Kernel Memory Model
How to Avoid Learning the Linux-Kernel Memory ModelHow to Avoid Learning the Linux-Kernel Memory Model
How to Avoid Learning the Linux-Kernel Memory Model
 
99.99% of Your Traces are Trash by Paige Cruz
99.99% of Your Traces are Trash by Paige Cruz99.99% of Your Traces are Trash by Paige Cruz
99.99% of Your Traces are Trash by Paige Cruz
 
Square's Lessons Learned from Implementing a Key-Value Store with Raft
Square's Lessons Learned from Implementing a Key-Value Store with RaftSquare's Lessons Learned from Implementing a Key-Value Store with Raft
Square's Lessons Learned from Implementing a Key-Value Store with Raft
 
Making Python 100x Faster with Less Than 100 Lines of Rust
Making Python 100x Faster with Less Than 100 Lines of RustMaking Python 100x Faster with Less Than 100 Lines of Rust
Making Python 100x Faster with Less Than 100 Lines of Rust
 
A Deep Dive Into Concurrent React by Matheus Albuquerque
A Deep Dive Into Concurrent React by Matheus AlbuquerqueA Deep Dive Into Concurrent React by Matheus Albuquerque
A Deep Dive Into Concurrent React by Matheus Albuquerque
 
The Latency Stack: Discovering Surprising Sources of Latency
The Latency Stack: Discovering Surprising Sources of LatencyThe Latency Stack: Discovering Surprising Sources of Latency
The Latency Stack: Discovering Surprising Sources of Latency
 

Recently uploaded

Cookies program to display the information though cookie creation
Cookies program to display the information though cookie creationCookies program to display the information though cookie creation
Cookies program to display the information though cookie creation
shanthidl1
 
TrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-In
TrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-InTrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-In
TrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-In
TrustArc
 
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
Kief Morris
 
WhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdf
WhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdfWhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdf
WhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdf
ArgaBisma
 
20240702 Présentation Plateforme GenAI.pdf
20240702 Présentation Plateforme GenAI.pdf20240702 Présentation Plateforme GenAI.pdf
20240702 Présentation Plateforme GenAI.pdf
Sally Laouacheria
 
Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024
BookNet Canada
 
Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...
BookNet Canada
 
WPRiders Company Presentation Slide Deck
WPRiders Company Presentation Slide DeckWPRiders Company Presentation Slide Deck
WPRiders Company Presentation Slide Deck
Lidia A.
 
Pigging Solutions Sustainability brochure.pdf
Pigging Solutions Sustainability brochure.pdfPigging Solutions Sustainability brochure.pdf
Pigging Solutions Sustainability brochure.pdf
Pigging Solutions
 
What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024
Stephanie Beckett
 
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - MydbopsScaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Mydbops
 
Quantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLMQuantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLM
Vijayananda Mohire
 
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Chris Swan
 
UiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs ConferenceUiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs Conference
UiPathCommunity
 
Advanced Techniques for Cyber Security Analysis and Anomaly Detection
Advanced Techniques for Cyber Security Analysis and Anomaly DetectionAdvanced Techniques for Cyber Security Analysis and Anomaly Detection
Advanced Techniques for Cyber Security Analysis and Anomaly Detection
Bert Blevins
 
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptxRPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
SynapseIndia
 
Best Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdfBest Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdf
Tatiana Al-Chueyr
 
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdfBT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
Neo4j
 
How RPA Help in the Transportation and Logistics Industry.pptx
How RPA Help in the Transportation and Logistics Industry.pptxHow RPA Help in the Transportation and Logistics Industry.pptx
How RPA Help in the Transportation and Logistics Industry.pptx
SynapseIndia
 
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdfINDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
jackson110191
 

Recently uploaded (20)

Cookies program to display the information though cookie creation
Cookies program to display the information though cookie creationCookies program to display the information though cookie creation
Cookies program to display the information though cookie creation
 
TrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-In
TrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-InTrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-In
TrustArc Webinar - 2024 Data Privacy Trends: A Mid-Year Check-In
 
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
 
WhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdf
WhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdfWhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdf
WhatsApp Image 2024-03-27 at 08.19.52_bfd93109.pdf
 
20240702 Présentation Plateforme GenAI.pdf
20240702 Présentation Plateforme GenAI.pdf20240702 Présentation Plateforme GenAI.pdf
20240702 Présentation Plateforme GenAI.pdf
 
Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024
 
Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...
 
WPRiders Company Presentation Slide Deck
WPRiders Company Presentation Slide DeckWPRiders Company Presentation Slide Deck
WPRiders Company Presentation Slide Deck
 
Pigging Solutions Sustainability brochure.pdf
Pigging Solutions Sustainability brochure.pdfPigging Solutions Sustainability brochure.pdf
Pigging Solutions Sustainability brochure.pdf
 
What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024
 
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - MydbopsScaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
 
Quantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLMQuantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLM
 
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
 
UiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs ConferenceUiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs Conference
 
Advanced Techniques for Cyber Security Analysis and Anomaly Detection
Advanced Techniques for Cyber Security Analysis and Anomaly DetectionAdvanced Techniques for Cyber Security Analysis and Anomaly Detection
Advanced Techniques for Cyber Security Analysis and Anomaly Detection
 
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptxRPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
 
Best Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdfBest Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdf
 
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdfBT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
 
How RPA Help in the Transportation and Logistics Industry.pptx
How RPA Help in the Transportation and Logistics Industry.pptxHow RPA Help in the Transportation and Logistics Industry.pptx
How RPA Help in the Transportation and Logistics Industry.pptx
 
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdfINDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
 

Scylla Summit 2017: Scylla on Samsung NVMe Z-SSDs

  • 1. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Scylla on Samsung Z-SSD Senior Performance Architect, Samsung Semiconductor Inc. Arash Rezaei
  • 2. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Arash Rezaei 2 Arash is a senior performance architect at Samsung Semiconductor Inc. His work has been primarily on performance evaluation and tuning of data center applications with a focus on storage.
  • 3. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Agenda 3 ▪ Z-SSD – Samsung’s ultra-low latency SSD ▪ Z-SSD use-cases ▪ Scylla on Z-SSD ▪ Summary
  • 4. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Introduction: Latency Matters 4 ▪ Threshold of patience is decreasing o Retrieving data from caches, databases, data stores o Business analytics MySQL/ NoSQL Databases Streaming Data Storage Layer Cache Layer Web Layer Analytics Latency
  • 5. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Samsung Z-SSD 5 ▪ Ultra-low latency flash storage device ▪ Provides 5.5x lower latency comparing to today’s leading NVMe SSD 14.7 80.3 0 10 20 30 40 50 60 70 80 90 Z-SSD PM1725a Latency [usec] 4K Random Read Latency 5.5x Lower Latency fio --rw=randread --bs=4k --direct=1 --iodepth=1 –numjobs=1 -- ioengine=libaio --norandommap
  • 6. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Samsung Z-SSD 6 DRAM Z-NAND New Controller 50%▼ Past LP4 70%▼ New CTRLPast x 1/10 ▼ Z-NANDPast ▪ Samsung’s Z-SSD is vertically optimized from DRAM to storage
  • 7. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company SZ985: Samsung Z-SSD 7 ▪ Sequential traffic o 3.2 GB/sec Read o 3.2 GB/sec Write ▪ Random traffic o 750K IOPs Reads o 170K IOPs Writes ▪ Capacity o 800GB
  • 8. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Z-SSD Minimizes Latency 8 ▪ High performance @ consistent low latency SSD Throughput Latency Sequential R/W Random R/W Random Read PM1725a 3.3/2.95 GB/s 800/160 KIOPS 90µs Z-SSD 3.2/3.2 GB/s 750/170 KIOPs 15µs Read-only workload (FIO) Mixed (80%/20% R/W) FIO
  • 9. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Z-SSD Use Cases
  • 10. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Z-SSD Benefits Latency-Sensitive Apps (1/2) 10 ▪ Z-SSD as swap space ▪ Memcached: distributed object caching system ▪ 3x speedup @ 1 3 the average latency under heavy swap load 0 20000 40000 60000 80000 100000 memcached Throughput[op/sec] Memcached Throughput PM1725a Z-SSD 0 200 400 600 800 1000 1200 1400 memcached ReadLatency[usec] Memcached Latency PM1725a Z-SSD 3x 𝟏 𝟑 memcached with 60GB dataset (45 million records, value size=1000B) DRAM capacity=32GB (less than half of the used memory space) Driven with YCSB, 90% read/ 10% update
  • 11. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Z-SSD Benefits Latency-Sensitive Apps (2/2) 11 ▪ Z-SSD as storage layer ▪ RocksDB is used as storage engine for DBs ▪ 2x throughput @ ½ the average latency 0 20000 40000 60000 80000 100000 120000 140000 RocksDB Throughput[op/sec] RocksDB Throughput PM1725a Z-SSD 2x 0 50 100 150 200 250 300 350 RocksDB AverateLatency[usec] RocksDB Average Latency PM1725a Z-SSD ½ db_bench read-while-writing workload
  • 12. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Scylla on Z-SSD
  • 13. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Methodology 13 ▪ Goal: o Compare Scylla in-memory vs. Scylla on Z-SSD ▪ Metric: o Max throughput under given tail-latency constraints ▪ Workload: o Cassandra-stress with: • Varying Read/Write ratio (100%/0% – 75/25% ) • Varying % of request served from memory (25% to 75%) o Database populated with 300GB data (payload 1KB)
  • 14. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company System Configuration 14 Item Description CPU 2 x Intel Xeon Gold 6154 @ 3.00GHz (total 36 cores per node) RAM 128GB SSD 2 x SZ985 800GB OS Ubuntu 16.04.2, Kernel 4.12.0 Scylla version 2.0
  • 15. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Throughput while Meeting Latency SLA (1/2) ▪ 100% read workload 15 Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
  • 16. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Throughput while Meeting Latency SLA (1/2) ▪ 100% read workload 16 Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
  • 17. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Throughput while Meeting Latency SLA (2/2) ▪ 75% read/25% write workload 17 Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
  • 18. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Throughput while Meeting Latency SLA (2/2) ▪ 75% read/25% write workload 18 Sub millisecond 95 percentile latency Sub 2 millisecond 99 percentile latency
  • 19. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company 19 ▪ Server viewpoint ▪ Latency under a fixed throughput ▪ 100% read workload (95 percentile latency) Software Overheads
  • 20. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company 20 ▪ Server viewpoint ▪ Latency under a fixed throughput ▪ 100% read workload (95 percentile latency) Software Overheads Device tail latency from FIO experiments
  • 21. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Scylla: DRAM, Z-SSD, or NVMe SSD? 21 ▪ For a given use-case requirement (throughput and DB size), what is the most cost-efficient cluster configuration? ▪ Optimize performance/$ for different use-cases
  • 22. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company Summary ▪ Samsung Z-SSD: ultra-low latency SSD device o 5.5x lower latency compared to NVMe SSD ▪ Z-SSD benefits latency-sensitive applications o RocksDB: 2x throughput @ ½ the average latency o Memcached (as swap): 3x speedup @ 1 3 the average latency ▪ Z-SSD for Scylla: o Helps reduce the performance gap between memory and storage • Delivers 44% – 23% of DRAM performance at significant lower cost – 50% – 75% of requests served from memory o Software improvements could further shrink the memory/storage gap o Opens new options for architecting a large cluster 22
  • 23. PRESENTATION TITLE ON ONE LINE AND ON TWO LINES First and last name Position, company THANK YOU a.rezaei@samsung.com Please stay in touch Any questions?