SlideShare una empresa de Scribd logo
1 de 17
Descargar para leer sin conexión
Jihyung Park, Hyuck Han and Sangyeun Cho
Memory Solutions Lab
Memory Business
Extending Main Memory with Flash –
the Optimized SWAP Approach
1. Introduction
2. Optimized SWAP
3. Evaluation
4. Future Work
5. Conclusion
Why extend main memory with flash?
• To overcome DRAM scaling limitations and offer large working memory
• To reduce total cost of ownership (acquisition and operation)
• Flash has no seek time
• Flash has faster latency than HDD
Two approaches toward memory extension
• Non-transparent approach: Application has to change
• Transparent approach: Application is NOT aware of the underlying flash
Introduction
Current swap algorithm is optimized for HDD
Paging for the Fast device
• Fast and Simple vs. Heavy and Accurate
Motivation
Swap entry search
• A new search algorithm
I/O path optimization
• Swap read-ahead
• I/O scheduler
• Swappiness
Swap device as backing store: Inclusive vs. Exclusive
• We adjust the swap entry free policy to enforce that the swap device
“includes” all swapped out pages
Optimized SWAP
Tree search
• “Bit tree”, no pointer, a node size is just one byte
• Fan-out degree is 8 (one bit is pointing a child node)
• 8-level tree covers multi-terabytes of swap space.
• Search cost: 2O(log N)
• Reduce swap structure size
– Roughly current swap mechanism vs. O-Swap = 10MB vs. 2MB (to support 32GB
swap space)
Optimized SWAP
0 2 4 61 3 5 7 8 9
Read-ahead
• No read-ahead (due to randomness)
• Note also that SSD has no seek time
I/O scheduler
• NOOP (due to randomness and fast response requirements)
• Bypass
Swappiness
• swappiness : 0
Swap entry reclaim policy
• Do not free swap entries as much as possible
Optimized SWAP
Evaluation - Memcached
System
CPU Xeon E5-2665 (HT disabled)
# Core 16
Network 10Gb Ethernet
SSD Samsung XS1715 (NVME)
Workload
YCSB
DB Size 30GB
Value Length 2048B
# memcached threads 64
# Clients 320
Get : Update 95% : 5%
Memory
SWAP OSWAP Full DRAM
DRAM 8GB
SSD Swap 32GB
DRAM 8GB
SSD Swap 32GB
DRAM 32GB
Evaluation - Memcached
0
2
4
6
8
10
12
14
SWAP OSWAP Full DRAM
Operationspersecond(x10,000)
Memcached (NVME, 10Gb Network)
Evaluation - Memcached
0
1
2
3
4
5
6
7
8
256us 512us 1024us 2ms 4ms 8ms 16ms 32ms 64ms 128ms 256ms 512ms
Operationspersecond(x1,000)
SWAP Performance by Latency Segment
< 1ms QoS
Evaluation - Memcached
0
5
10
15
20
25
256us 512us 1024us 2ms 4ms 8ms 16ms 32ms 64ms 128ms 256ms 512ms
Operationspersecond(x1,000)
OSWAP Performance by Latency Segment
< 1ms QoS
Evaluation - Memcached
0
2
4
6
8
10
12
256us 512us 1024us 2ms 4ms 8ms 16ms 32ms 64ms 128ms
Operationspersecond(x10,000)
Full DRAM Performance by Latency Segment
< 1ms QoS
Evaluation - Linkbench
System
CPU Xeon E5-2665 (HT disabled)
# Core 16
Network 10Gb Ethernet
SSD Samsung XS1715 (NVME)
Workload
Linkbench
DB Size 30GB
# Clients 400
Memory
SWAP OSWAP Full DRAM
DRAM 8GB
SSD Swap 32GB
DRAM 8GB
SSD Swap 32GB
DRAM 32GB
Evaluation - Linkbench
0
2
4
6
8
10
12
14
SWAP OSWAP Full DRAM
Requestspersecond(x1,000)
Linkbench
Rack scale architecture
High performance memory + High capacity memory
Future Work
CPUs
DRAM
DRAM
DRAM
Compute
PCIe <-> Ctrl Ctrl
Memory
Memory
Memorycable
Memory Device
Cost-effective memory capacity
Exploit flash memory transparently
Conclusion
Nvmw 2014  extending main memory with flash-the optimized swap approach

Más contenido relacionado

La actualidad más candente

Scaling wordpress for high traffic
Scaling wordpress for high trafficScaling wordpress for high traffic
Scaling wordpress for high traffic
Roshan Bhattarai
 
Roshan Bhattarai: Scaling WordPress for high traffic sites
Roshan Bhattarai: Scaling WordPress for high traffic sitesRoshan Bhattarai: Scaling WordPress for high traffic sites
Roshan Bhattarai: Scaling WordPress for high traffic sites
wpnepal
 
Nimble-Storage-AFA-Datasheet
Nimble-Storage-AFA-DatasheetNimble-Storage-AFA-Datasheet
Nimble-Storage-AFA-Datasheet
Mike Finnegan
 
Using flash on the server side
Using flash on the server sideUsing flash on the server side
Using flash on the server side
Howard Marks
 

La actualidad más candente (19)

dh-slides-perf.ppt
dh-slides-perf.pptdh-slides-perf.ppt
dh-slides-perf.ppt
 
Hbase
HbaseHbase
Hbase
 
Hbase
HbaseHbase
Hbase
 
Redis on NVMe SSD - Zvika Guz, Samsung
 Redis on NVMe SSD - Zvika Guz, Samsung Redis on NVMe SSD - Zvika Guz, Samsung
Redis on NVMe SSD - Zvika Guz, Samsung
 
Brief of Caching
Brief of CachingBrief of Caching
Brief of Caching
 
Brief of Caching - Rafiul Islam
Brief of Caching - Rafiul IslamBrief of Caching - Rafiul Islam
Brief of Caching - Rafiul Islam
 
MySQL Performance - SydPHP October 2011
MySQL Performance - SydPHP October 2011MySQL Performance - SydPHP October 2011
MySQL Performance - SydPHP October 2011
 
Server side caching Vs other alternatives
Server side caching Vs other alternativesServer side caching Vs other alternatives
Server side caching Vs other alternatives
 
Breaking the Sound Barrier with Persistent Memory
Breaking the Sound Barrier with Persistent Memory Breaking the Sound Barrier with Persistent Memory
Breaking the Sound Barrier with Persistent Memory
 
Scaling wordpress for high traffic
Scaling wordpress for high trafficScaling wordpress for high traffic
Scaling wordpress for high traffic
 
Roshan Bhattarai: Scaling WordPress for high traffic sites
Roshan Bhattarai: Scaling WordPress for high traffic sitesRoshan Bhattarai: Scaling WordPress for high traffic sites
Roshan Bhattarai: Scaling WordPress for high traffic sites
 
Nimble-Storage-AFA-Datasheet
Nimble-Storage-AFA-DatasheetNimble-Storage-AFA-Datasheet
Nimble-Storage-AFA-Datasheet
 
Webinar: Introduction to MongoDB 3.0
Webinar: Introduction to MongoDB 3.0Webinar: Introduction to MongoDB 3.0
Webinar: Introduction to MongoDB 3.0
 
SVC / Storwize analysis cost effective storage planning (use case)
SVC / Storwize analysis cost effective storage planning (use case)SVC / Storwize analysis cost effective storage planning (use case)
SVC / Storwize analysis cost effective storage planning (use case)
 
WiredTiger MongoDB Integration
WiredTiger MongoDB Integration WiredTiger MongoDB Integration
WiredTiger MongoDB Integration
 
Ceph Day Amsterdam 2015 - Ceph backing the first Government Cloud in the Neth...
Ceph Day Amsterdam 2015 - Ceph backing the first Government Cloud in the Neth...Ceph Day Amsterdam 2015 - Ceph backing the first Government Cloud in the Neth...
Ceph Day Amsterdam 2015 - Ceph backing the first Government Cloud in the Neth...
 
Hybrid Storage Pools (Now with the benefit of hindsight!)
Hybrid Storage Pools (Now with the benefit of hindsight!)Hybrid Storage Pools (Now with the benefit of hindsight!)
Hybrid Storage Pools (Now with the benefit of hindsight!)
 
Using flash on the server side
Using flash on the server sideUsing flash on the server side
Using flash on the server side
 
Off-heaping the Apache HBase Read Path
Off-heaping the Apache HBase Read Path Off-heaping the Apache HBase Read Path
Off-heaping the Apache HBase Read Path
 

Destacado

Dram row-hammer kim-talk_isca14
Dram row-hammer kim-talk_isca14Dram row-hammer kim-talk_isca14
Dram row-hammer kim-talk_isca14
Benoit Hudzia
 
Enhancing Live Migration Process for CPU and/or memory intensive VMs running...
Enhancing Live Migration Process for CPU and/or  memory intensive VMs running...Enhancing Live Migration Process for CPU and/or  memory intensive VMs running...
Enhancing Live Migration Process for CPU and/or memory intensive VMs running...
Benoit Hudzia
 
Flash Memory Ted 111
Flash Memory Ted 111Flash Memory Ted 111
Flash Memory Ted 111
mil1375
 
A new era in embedded flash memory, Anobit presentation fms 2011 -
A new era in embedded flash memory, Anobit presentation fms 2011 -A new era in embedded flash memory, Anobit presentation fms 2011 -
A new era in embedded flash memory, Anobit presentation fms 2011 -
Anobit
 

Destacado (17)

Lego Cloud SAP Virtualization Week 2012
Lego Cloud SAP Virtualization Week 2012Lego Cloud SAP Virtualization Week 2012
Lego Cloud SAP Virtualization Week 2012
 
Hecatonchire kvm forum_2012_benoit_hudzia
Hecatonchire kvm forum_2012_benoit_hudziaHecatonchire kvm forum_2012_benoit_hudzia
Hecatonchire kvm forum_2012_benoit_hudzia
 
Dram row-hammer kim-talk_isca14
Dram row-hammer kim-talk_isca14Dram row-hammer kim-talk_isca14
Dram row-hammer kim-talk_isca14
 
Enhancing Live Migration Process for CPU and/or memory intensive VMs running...
Enhancing Live Migration Process for CPU and/or  memory intensive VMs running...Enhancing Live Migration Process for CPU and/or  memory intensive VMs running...
Enhancing Live Migration Process for CPU and/or memory intensive VMs running...
 
Spansion FL-S Serial NOR Flash Memory
Spansion FL-S Serial NOR Flash MemorySpansion FL-S Serial NOR Flash Memory
Spansion FL-S Serial NOR Flash Memory
 
Flash memory
Flash memoryFlash memory
Flash memory
 
Hush…tell you something novel about flash memory
Hush…tell you something novel about flash memoryHush…tell you something novel about flash memory
Hush…tell you something novel about flash memory
 
Flash Memory Ted 111
Flash Memory Ted 111Flash Memory Ted 111
Flash Memory Ted 111
 
"Achieving Flash Memory's Full Potential" @ Flash Memory Summit 2012
"Achieving Flash Memory's Full Potential" @ Flash Memory Summit 2012"Achieving Flash Memory's Full Potential" @ Flash Memory Summit 2012
"Achieving Flash Memory's Full Potential" @ Flash Memory Summit 2012
 
Flash memory summit 2015 gary lyng session 301-a
Flash memory summit 2015   gary lyng session 301-aFlash memory summit 2015   gary lyng session 301-a
Flash memory summit 2015 gary lyng session 301-a
 
Flash memory device inside
Flash memory device insideFlash memory device inside
Flash memory device inside
 
Flash memory as a hard drive replacement
Flash  memory  as a  hard  drive replacementFlash  memory  as a  hard  drive replacement
Flash memory as a hard drive replacement
 
A new era in embedded flash memory, Anobit presentation fms 2011 -
A new era in embedded flash memory, Anobit presentation fms 2011 -A new era in embedded flash memory, Anobit presentation fms 2011 -
A new era in embedded flash memory, Anobit presentation fms 2011 -
 
Persistent memory
Persistent memoryPersistent memory
Persistent memory
 
Nand flash memory
Nand flash memoryNand flash memory
Nand flash memory
 
Flash Memory Storage
Flash Memory StorageFlash Memory Storage
Flash Memory Storage
 
Hana Memory Scale out using the hecatonchire Project
Hana Memory Scale out using the hecatonchire ProjectHana Memory Scale out using the hecatonchire Project
Hana Memory Scale out using the hecatonchire Project
 

Similar a Nvmw 2014 extending main memory with flash-the optimized swap approach

Colvin exadata mistakes_ioug_2014
Colvin exadata mistakes_ioug_2014Colvin exadata mistakes_ioug_2014
Colvin exadata mistakes_ioug_2014
marvin herrera
 
Handling Massive Writes
Handling Massive WritesHandling Massive Writes
Handling Massive Writes
Liran Zelkha
 
In-memory Caching in HDFS: Lower Latency, Same Great Taste
In-memory Caching in HDFS: Lower Latency, Same Great TasteIn-memory Caching in HDFS: Lower Latency, Same Great Taste
In-memory Caching in HDFS: Lower Latency, Same Great Taste
DataWorks Summit
 
Nimble storage investor_deck_public
Nimble storage investor_deck_publicNimble storage investor_deck_public
Nimble storage investor_deck_public
Sequoia Capital
 

Similar a Nvmw 2014 extending main memory with flash-the optimized swap approach (20)

Colvin exadata mistakes_ioug_2014
Colvin exadata mistakes_ioug_2014Colvin exadata mistakes_ioug_2014
Colvin exadata mistakes_ioug_2014
 
Accelerating HBase with NVMe and Bucket Cache
Accelerating HBase with NVMe and Bucket CacheAccelerating HBase with NVMe and Bucket Cache
Accelerating HBase with NVMe and Bucket Cache
 
In-memory Data Management Trends & Techniques
In-memory Data Management Trends & TechniquesIn-memory Data Management Trends & Techniques
In-memory Data Management Trends & Techniques
 
Ceph Community Talk on High-Performance Solid Sate Ceph
Ceph Community Talk on High-Performance Solid Sate Ceph Ceph Community Talk on High-Performance Solid Sate Ceph
Ceph Community Talk on High-Performance Solid Sate Ceph
 
Using Big Data techniques to query and store OpenStreetMap data. Stephen Knox...
Using Big Data techniques to query and store OpenStreetMap data. Stephen Knox...Using Big Data techniques to query and store OpenStreetMap data. Stephen Knox...
Using Big Data techniques to query and store OpenStreetMap data. Stephen Knox...
 
FlashSQL 소개 & TechTalk
FlashSQL 소개 & TechTalkFlashSQL 소개 & TechTalk
FlashSQL 소개 & TechTalk
 
#GeodeSummit - Off-Heap Storage Current and Future Design
#GeodeSummit - Off-Heap Storage Current and Future Design#GeodeSummit - Off-Heap Storage Current and Future Design
#GeodeSummit - Off-Heap Storage Current and Future Design
 
Taking Splunk to the Next Level - Architecture Breakout Session
Taking Splunk to the Next Level - Architecture Breakout SessionTaking Splunk to the Next Level - Architecture Breakout Session
Taking Splunk to the Next Level - Architecture Breakout Session
 
Apache Geode Offheap Storage
Apache Geode Offheap StorageApache Geode Offheap Storage
Apache Geode Offheap Storage
 
Red Hat Ceph Storage Acceleration Utilizing Flash Technology
Red Hat Ceph Storage Acceleration Utilizing Flash Technology Red Hat Ceph Storage Acceleration Utilizing Flash Technology
Red Hat Ceph Storage Acceleration Utilizing Flash Technology
 
Make Oracle scream with Flash Storage - Kaminario
Make Oracle scream with Flash Storage - KaminarioMake Oracle scream with Flash Storage - Kaminario
Make Oracle scream with Flash Storage - Kaminario
 
JSSUG: SQL Sever Performance Tuning
JSSUG: SQL Sever Performance TuningJSSUG: SQL Sever Performance Tuning
JSSUG: SQL Sever Performance Tuning
 
Handling Massive Writes
Handling Massive WritesHandling Massive Writes
Handling Massive Writes
 
In-memory Caching in HDFS: Lower Latency, Same Great Taste
In-memory Caching in HDFS: Lower Latency, Same Great TasteIn-memory Caching in HDFS: Lower Latency, Same Great Taste
In-memory Caching in HDFS: Lower Latency, Same Great Taste
 
Taking Splunk to the Next Level - Architecture
Taking Splunk to the Next Level - ArchitectureTaking Splunk to the Next Level - Architecture
Taking Splunk to the Next Level - Architecture
 
Nimble storage investor_deck_public
Nimble storage investor_deck_publicNimble storage investor_deck_public
Nimble storage investor_deck_public
 
The state of SQL-on-Hadoop in the Cloud
The state of SQL-on-Hadoop in the CloudThe state of SQL-on-Hadoop in the Cloud
The state of SQL-on-Hadoop in the Cloud
 
IMCSummit 2015 - Day 1 Developer Session - The Science and Engineering Behind...
IMCSummit 2015 - Day 1 Developer Session - The Science and Engineering Behind...IMCSummit 2015 - Day 1 Developer Session - The Science and Engineering Behind...
IMCSummit 2015 - Day 1 Developer Session - The Science and Engineering Behind...
 
Ceph Day San Jose - Red Hat Storage Acceleration Utlizing Flash Technology
Ceph Day San Jose - Red Hat Storage Acceleration Utlizing Flash TechnologyCeph Day San Jose - Red Hat Storage Acceleration Utlizing Flash Technology
Ceph Day San Jose - Red Hat Storage Acceleration Utlizing Flash Technology
 
Optimizing Latency-sensitive queries for Presto at Facebook: A Collaboration ...
Optimizing Latency-sensitive queries for Presto at Facebook: A Collaboration ...Optimizing Latency-sensitive queries for Presto at Facebook: A Collaboration ...
Optimizing Latency-sensitive queries for Presto at Facebook: A Collaboration ...
 

Último

Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Safe Software
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Safe Software
 

Último (20)

Mcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot Model
Mcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot ModelMcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot Model
Mcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot Model
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 
Elevate Developer Efficiency & build GenAI Application with Amazon Q​
Elevate Developer Efficiency & build GenAI Application with Amazon Q​Elevate Developer Efficiency & build GenAI Application with Amazon Q​
Elevate Developer Efficiency & build GenAI Application with Amazon Q​
 
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost SavingRepurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf
 
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
 
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWEREMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
 
Understanding the FAA Part 107 License ..
Understanding the FAA Part 107 License ..Understanding the FAA Part 107 License ..
Understanding the FAA Part 107 License ..
 
Artificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyArtificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : Uncertainty
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a Fresher
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
DBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor PresentationDBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor Presentation
 
Six Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal OntologySix Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal Ontology
 
DEV meet-up UiPath Document Understanding May 7 2024 Amsterdam
DEV meet-up UiPath Document Understanding May 7 2024 AmsterdamDEV meet-up UiPath Document Understanding May 7 2024 Amsterdam
DEV meet-up UiPath Document Understanding May 7 2024 Amsterdam
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 

Nvmw 2014 extending main memory with flash-the optimized swap approach

  • 1. Jihyung Park, Hyuck Han and Sangyeun Cho Memory Solutions Lab Memory Business Extending Main Memory with Flash – the Optimized SWAP Approach
  • 2. 1. Introduction 2. Optimized SWAP 3. Evaluation 4. Future Work 5. Conclusion
  • 3. Why extend main memory with flash? • To overcome DRAM scaling limitations and offer large working memory • To reduce total cost of ownership (acquisition and operation) • Flash has no seek time • Flash has faster latency than HDD Two approaches toward memory extension • Non-transparent approach: Application has to change • Transparent approach: Application is NOT aware of the underlying flash Introduction
  • 4. Current swap algorithm is optimized for HDD Paging for the Fast device • Fast and Simple vs. Heavy and Accurate Motivation
  • 5. Swap entry search • A new search algorithm I/O path optimization • Swap read-ahead • I/O scheduler • Swappiness Swap device as backing store: Inclusive vs. Exclusive • We adjust the swap entry free policy to enforce that the swap device “includes” all swapped out pages Optimized SWAP
  • 6. Tree search • “Bit tree”, no pointer, a node size is just one byte • Fan-out degree is 8 (one bit is pointing a child node) • 8-level tree covers multi-terabytes of swap space. • Search cost: 2O(log N) • Reduce swap structure size – Roughly current swap mechanism vs. O-Swap = 10MB vs. 2MB (to support 32GB swap space) Optimized SWAP 0 2 4 61 3 5 7 8 9
  • 7. Read-ahead • No read-ahead (due to randomness) • Note also that SSD has no seek time I/O scheduler • NOOP (due to randomness and fast response requirements) • Bypass Swappiness • swappiness : 0 Swap entry reclaim policy • Do not free swap entries as much as possible Optimized SWAP
  • 8. Evaluation - Memcached System CPU Xeon E5-2665 (HT disabled) # Core 16 Network 10Gb Ethernet SSD Samsung XS1715 (NVME) Workload YCSB DB Size 30GB Value Length 2048B # memcached threads 64 # Clients 320 Get : Update 95% : 5% Memory SWAP OSWAP Full DRAM DRAM 8GB SSD Swap 32GB DRAM 8GB SSD Swap 32GB DRAM 32GB
  • 9. Evaluation - Memcached 0 2 4 6 8 10 12 14 SWAP OSWAP Full DRAM Operationspersecond(x10,000) Memcached (NVME, 10Gb Network)
  • 10. Evaluation - Memcached 0 1 2 3 4 5 6 7 8 256us 512us 1024us 2ms 4ms 8ms 16ms 32ms 64ms 128ms 256ms 512ms Operationspersecond(x1,000) SWAP Performance by Latency Segment < 1ms QoS
  • 11. Evaluation - Memcached 0 5 10 15 20 25 256us 512us 1024us 2ms 4ms 8ms 16ms 32ms 64ms 128ms 256ms 512ms Operationspersecond(x1,000) OSWAP Performance by Latency Segment < 1ms QoS
  • 12. Evaluation - Memcached 0 2 4 6 8 10 12 256us 512us 1024us 2ms 4ms 8ms 16ms 32ms 64ms 128ms Operationspersecond(x10,000) Full DRAM Performance by Latency Segment < 1ms QoS
  • 13. Evaluation - Linkbench System CPU Xeon E5-2665 (HT disabled) # Core 16 Network 10Gb Ethernet SSD Samsung XS1715 (NVME) Workload Linkbench DB Size 30GB # Clients 400 Memory SWAP OSWAP Full DRAM DRAM 8GB SSD Swap 32GB DRAM 8GB SSD Swap 32GB DRAM 32GB
  • 14. Evaluation - Linkbench 0 2 4 6 8 10 12 14 SWAP OSWAP Full DRAM Requestspersecond(x1,000) Linkbench
  • 15. Rack scale architecture High performance memory + High capacity memory Future Work CPUs DRAM DRAM DRAM Compute PCIe <-> Ctrl Ctrl Memory Memory Memorycable Memory Device
  • 16. Cost-effective memory capacity Exploit flash memory transparently Conclusion