M|18 Intel and MariaDB: Strategic Collaboration to Enhance MariaDB Functionality, Performance and TCO
•Download as PPTX, PDF•
3 likes•1,442 views
M|18 Intel and MariaDB: Strategic Collaboration to Enhance MariaDB Functionality, Performance and TCO
Recommended
More Related Content
What's hot
What's hot (20)
Similar to M|18 Intel and MariaDB: Strategic Collaboration to Enhance MariaDB Functionality, Performance and TCO
Similar to M|18 Intel and MariaDB: Strategic Collaboration to Enhance MariaDB Functionality, Performance and TCO (20)
More from MariaDB plc
More from MariaDB plc (20)
Recently uploaded
Recently uploaded (20)
M|18 Intel and MariaDB: Strategic Collaboration to Enhance MariaDB Functionality, Performance and TCO
- 1. Strategic collaboration to enhance MariaDB functionality, performance & TCO
- 2. Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. Check with your system manufacturer or retailer or learn more at intel.com. No computer system can be absolutely secure. Tests document performance of components on a particular test, in specific systems. Differences in hardware, software, or configuration will affect actual performance. Consult other sources of information to evaluate performance as you consider your purchase. For more complete information about performance and benchmark results, visit http://www.intel.com/performance. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance. Cost reduction scenarios described are intended as examples of how a given Intel-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction. Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. © 2018 Intel Corporation. Intel, the Intel logo, and Intel Xeon are trademarks of Intel Corporation in the U.S. and/or other countries. *Other names and brands may be claimed as property of others. 2 Notices & Disclaimers
- 3. • Collaborating to transform MariaDB transaction performance and cost • Enabling reliable, real-time web-scale applications through distributed log capabilities • Accelerating MariaDB with Intel® FPGAs 3 Agenda
- 5. Cloud & DATA Center Things & Devices MEM ORY FPGA Cloud and Data Center, AI, SD Infrastructure, Big Data Analytics Manufacturing, Memory, FPGA, Software, Security Intel Capital leads $20 million investment to grow MariaDB Open source database gets funding to boost adoption and develop product range Intel® Xeon® Processor Scalable Family 3D XPointTM Intel® OptaneTM SSD DC P4800X Intel® Stratix® 10 FPGA Virtuous Cycle of Growth
- 6. Intel’s Software Group: MariaDB Team 6 Shanghai Intel DCG database performance lab (Intel® Xeon®) Moscow MariaDB database performance Early Access lab London Database performance tools & customer engagements Munich MariaDB - Intel Global Relationship Management
- 7. 7 Working together to optimize the combination of: • MariaDB* and MyRocks* storage engine • Intel® Xeon® Scalable Platform • Intel® Optane™ SSDs Co-Marketing Collaborations MariaDB Intel Collaboration Early access to Intel pre-production H/W GoToMarket support and customer Trusted Advisors Joint performance analysis and tuning S/W development tools Full tech. support Long-term development project collaboration MyRocks
- 8. delivers 1.65x average performance boost over prior Generation1 Intel® Xeon® Scalable platform http://www.intel.com/performance Perform ance Agilit y Securi tyPervasive through compute, storage and network Pervasive data security with no performance overhead Rapid service delivery The Foundation of Data Center Innovation: Agile & Trusted Infrastructure
- 9. MariaDB OLTP Performance on Intel® Xeon® Scalable PLATFORM 9 0 175000 350000 525000 700000 Xeon E5-2697v3 Xeon E5-2699v4 Xeon Platinum 8180 NOPM 1.26X 1.34X 1.7X HammerDB test provides a complex OLTP workload with write/write locking contention to simulate real-world applications and compare performance across database platforms http://www.intel.com/performance MariaDB OLTP Performance on Intel® Xeon® Scalable PLATFORM MariaDB 10.2 InnoDB OLTP Workload
- 10. 10 0. 0.3 0.6 0.9 1.2 1.5 MariaDB Commercial $ 0 0.35 0.7 1.05 1.4 MariaDB Commercial $ 0. 0.25 0.5 0.75 1. MariaDB Commercial $ 32X 41X 42X • Software Total Cost of Ownership (TCO) over 3 years MariaDB and commercial database • ‘Cost per Transaction’ means cost for the same unit of work • MariaDB becomes more compelling with each CPU generation $0.038 $1.20 $1.23 $0.030 $0.92 $0.022 http://www.intel.com/performance Lower Cost Lower Cost Lower Cost MariaDB OLTP Cost on Intel® Xeon® Scalable Platform 2014 Cost per Transaction E5-2697v3 2016 Cost per Transaction E5-2699v4 2017 Cost per Transaction Platinum 8180
- 11. A Glimpse Inside the Intel® Xeon® Scalable platform SSD Intel® Optane™ SSD DC P4800X Complementary Intel® FPGA Workload optimized frameworks Integrated Options Fabric Intel® Omni-Path Architecture Networking Intel® Ethernet Accelerators Intel® QuickAssist Intel® AVX-512 Advancing virtually every aspect: Brand New core, cache, on-die interconnects, memory controller & more performa nce Agilitysecurity
- 12. 12 Combining the attributes of memory and storage with low latency, high endurance, outstanding QoS and high throughput+ Intel® optane™ SSD DC P4800X series is delivering on data center needs Enabling a new data tier to accelerate applications for fast caching & storage and expanded memory pools, to increase scale per server and reduce transaction cost Bringing synergy with Intel® Xeon® processors to enable bigger and more affordable datasets to gain new insights from larger memory pools
- 13. 13 MySQL* 5.7 Sysbench internal testing Memory DRAM 256GB DRAM 256GB DRAM 256GB CPU Intel® Xeon® v4 (44 core) Intel® Xeon™ Intel® Xeon® v4 (44 core) Intel® Xeon™ Intel® Xeon® v4 (44 core) Intel® Xeon™ Storage Intel® SSD DC S3520 Intel® SSD DC P4500 Intel® Optane™ SSD Transactions Per Second 17,235 5,834 3,535 TPS1 4x better up to 8x better up to Latency1 Intel® Optane™ ssd vs. “good” solution P99 Latency 12 ms 51 ms 100 ms 1.System configuration:Server Intel® Server System R2208WT2YS, 2x Intel® Xeon® E5 2699v4, 384 GB DDR4 DRAM, boot drive- 1x Intel® SSD DC S3710 Series (400 GB), database drives- 1x Intel® SSD DC P3700 Series (400 GB) and 1x Intel® Optane™ SSD DC P4800X Series (140 GB prototype),CentOS 7.2, MySQL Server 5.7.14, Sysbench 0.5 configuredfor 70/30 Read/WriteOLTP transactionsplit using a 100GB database.Cost per transactiondetermined by totalMSRP for each configurationdivided by the transactionsper second. Estimatedresultswereobtainedpriortoimplementationofrecentsoftwarepatchesandfirmwareupdatesintendedtoaddressexploitsreferredtoas"Spectre"and "Meltdown".Implementationoftheseupdatesmaymaketheseresultsinapplicabletoyourdeviceorsystem. *Other names and brands names may be claimed as thepropertyof others
- 14. 14 Xeon E5-2699v4 Xeon Platinum 8180 1.78X MariaDB*: Sysbench OLTP workload. OS: Ubuntu Server 17.04 x64. Sysbench 0.4.12, MariaDB 10.2.9 GIT snapshot 2017-09-12. Testing by Intel, September 2017. BASELINE: 2S Intel® Xeon® E5-2699 v4, 2.2GHz, 22 cores, turbo and HT on, BIOS 63.R00, 64GB total memory, 1600 MT/s / DDR4 LRDIMM, 1TB SATA HDD, Intel® SSD DC P3700 Series 2TB . NEW: 2S Intel® Xeon® Platinum 8180 processor, 2.5 GHz, 28 cores, turbo and HT on, BIOS 1.00.0412, 192GB total memory, 12 slots / 16GB / 2666 MT/s / DDR4 LRDIMM, 1 TB SATA HDD, Intel® SSD DC P3700 Series 2TB. http://www.intel.com/performance “MyRocks is a storage engine that adds the RocksDB LSM flash storage-optimized database to MariaDB. Even while under development, it shows good performance and scalability -- close to mature storage engines like InnoDB. Running our pre-release MariaDB 10.2.9 with a built-in MyRocks engine under a Sysbench RO multi-table test on the newest Intel® Xeon® Platinum 8180 processors, we saw up to 1.78x more throughput with a 1.18x reduction of average response time vs. previous-gen Intel® Xeon® processor E5-2699 v4 performance.” Michael Widenius, CTO, MariaDB Corporation Ab MYROCKS* NEXT-GENERATION STORAGE ENGINE ON Intel® Xeon® Scalable PlatformMariaDB 10.2.9 Sysbench OLTP Workload
- 15. 15 Enabling reliable, real-time web-scale applications through distributed log capabilities
- 16. Database Engine Shared Storage Clustering Software Oracle RAC Oracle Automatic Storage Management Oracle Clusterware* The Stack Hardware x86 servers interconnected via Infiniband* (or Ethernet) Oracle RAC* as the “Gold Standard”
- 17. 17 1.AWS Relational Database Service (RDS) 2.AWS Aurora (RDS) “Data Center Networks Are in My Way!” Using AWS’s RDS* vis-à-vis their RDS/Aurora* Log-Structured Storage as the example
- 18. 18 Matsunobu, "MyRocks: Space and write optimized OLTP database at Facebook," 2016 https://atscaleconference.com/videos/myrocks-space-and-write-optimized-oltp-database-at-facebook/ Facebook’s MyRocks* Comes Out from Behind the Curtain Major FEATURES in MyRocks Similar Feature Sets as InnoDB* TRANSAC TIONS ONLINE BACKUP• Atomicity • Non-locking consistent reads • Read committed • Repeatable read • Crash-safe slave and master • Logical backup by mysqldump • Binary backup by myrocks_hotbackup
- 19. 19 Customer “Pain Points” The Downside of Becoming a Data Company Companies that have pivoted to AI and analytics in order to monetize their data are experiencing accelerating data growth rates Disaggregating Storage This data growth challenge makes disaggregating storage from compute attractive because the company can scale their storage capacity to match their data growth, independent of compute. Cloud-Like, Open-Source, and AI/Analytics-Ready Many of these companies are big Oracle RAC*/Exadata* users. They expressed a desire to move off this platform to something more cloud-like, open source, and readily integrated into their AI/analytics investments. 1 2 3
- 20. 20 Physical Cluster ToR Switch Spine Switch JBOF Storage Servers Compute Servers 100GbsLinks 50GbsLinks Resource Mgmt Service Mgmt Services A Solution to Address These Pain Points CLUSTER SOFTWARE
- 21. 21 Revisiting Cluster Computing in the Era of Cloud Rack-Scale Design and Disaggregated Storage 1.Rack-Centric, Physical Cluster 2.Cluster-Wide Service Management (k8s) 3.Database-as-a- Service on Log- Structured Storage website1 website2 ELB Service (Load balancer) Pod ingress controller Ingress Service (Node Port) Service (Node Port) … Service (Node Port) Pod (website1) Pod (website2) Pod (default backend) Kubernetes
- 22. • Support for the persistence requirements of distributed databases • Fast recovery and cloning through an automated recovery process - Replicated log service • Little or no interruptions to the transaction processing - High-throughput and low-latency - Durability, replication and strong consistency - Optimizes both performance and TCO for transaction-centric workloads 22 Advantages of Log- Structured Storage
- 23. 23 A Stack Comparison A Stack Comparison Enterprise (Oracle) Cloud (Amazon Aurora*) Intel Solution (MariaDB*-based Solution) Database Engine Proprietary (Oracle RAC*) Open Source (MySQL, Postgres) Open Source MariaDB with the MaxScale Layer-7 Proxy Shared Storage Proprietary Oracle Automatic Storage Management (ASM) with POSIX File System interface Open Source +Proprietary (Distributed Log, S3) Open Source MyRocks* compiled with Rockset’s RocksDB-Cloud* library Clustering Software Proprietary (Oracle Clusterware*) Proprietary AWS RDS* Open Source Kubernetes* Hardware Proprietary x86 servers interconnected via Infiniband* (or Ethernet) Commodity x86 servers interconnected via Ethernet Fabric Commodity x86 servers interconnected via 100Gbs Ethernet Fabric
- 24. Flush sst file to local SSD writes memtable cache Persistent read cache on SSD Flush sst file to cloud storage Cloud Storage Bucket Cloud Application RocksDB-Cloud* block cache reads Queries Updates Cloud Log Storage, Kafka* Topic 24 Provisioning a Database Instance WAL
- 25. read write Cloned Server RocksDB-Cloud read Cloud Bucket A Server RocksDB-Cloud* Clones are read-only replicas 25 fast cloning Queries Updates Queries Cloud Bucket B write read Cloud Log Storage, Kafka* WAL writes WAL tailer
- 26. • Uses MyRocks w/Rockset’s RocksDB-Cloud library • Requires Support for Lossless-SemiSync Protocol and GTIDs in the MaxLog Binlog Server (“Log-Tailer”) • Need a “Lambda” function to kickoff “Switchover” • Need Automation for carrying out Switchover • Can use either InnoDB* or MyRocks* w/RocksDB* library • Requires Support for Lossless-SemiSync Protocol and GTIDs in the Slaves • Need a “Lambda” function to kickoff “Switchover” • Need Automation for carrying out Switchover 26 Time0 Time1 Time0 Time1 Database Instance within a Single Cluster1.Master-Slave (Shared-Nothing) 2.Master-Only (Shared)
- 27. 27 What’s Next? Enabling the use of Nonvolatile Memory (NVM)
- 28. 28 Improve real-time data analytics and data warehousing TCO with Intel® FPGAs
- 29. Intel® Big Data Analytics Frameworks Accelerate innovation in big data analytics with frameworks built on software-defined Infrastructure with open-standard building blocks. Intel® Frameworks & Libraries Integrated with FPGAs Run unmodified customer applications, use runtime orchestration with both Intel® Xeon® processor and FPGA support, and leverage end-to-end virtualization and security. Accelerate Relational, NoSQL and Unstructured FPGA data access, networking and algorithm acceleration options with a single FPGA for highly structured, semi- structured, and unstructured data for better TCO, flexibility and future proofing. Analytics Landscape and Scaling Accelerate Big Data Analytics with Existing Interfaces and FPGAs 1 2 3
- 31. Microsoft Scale-Out FPGA Multi-Function Accelerator • “Diversity of cloud workloads and … rapid … change” (weekly or monthly) – Search, SmartNIC, Machine Learning, Encrypt, Compress, Big Data Analytics… • Bing Search: 2X server level perf, 29% latency reduction • Networking Virtualization: 10X latency improvement • Machine Learning: Stratix 10 capable of 90 TFLOPs 8-bit floating-point Single FPGA Algorithm, Networking & Data Access Acceleration
- 32. 32 Data Network Streaming Data Integrate to Intel® Frameworks and APIs – Run unmodified customer applications – Orchestration run-time advantage: Xeon® or FPGA – End-to-end security and virtualization framework Moderate Acceleration is Common – PCIe* look-aside acceleration (two data copies) Significant Acceleration requires FPGA – Multifunction and inline with a single FPGA FPGAs Offer Unique Value for Analytics & Streaming Offloads algorithm, networking and data access processing Single Multifunction Accelerator
- 34. 34 Acceleration Overview • 20X+ single table inserts/s for real time data analytics - With modest tuning, 15M INSERT/s1 • 2X+ optimized queries for data warehousing - Using industry standard TPC-DS benchmark • 3X+ storage compression - Data and tables managed by Swarm64* Swarm64 Relational Database Acceleration Two Workloads: Traditional Data Warehousing, Real-Time Data Analytics Database acceleration with a plugin
- 35. 35 Swarm64 Relational Database Acceleration Scale-Up Data Warehousing, Real-Time Data Analytics, and Storage Compression Database acceleration with a plugin Overview • No customer application change - Storage engine plugin - Query Engine accelerates INSERT, SELECT, … • Optimized indexing • More I/O bandwidth, memory depth from compression Significant Acceleration √ Data access acceleration √ Compression, filtering, replication … √ Memory-mapped acceleration, data cache √ “Optimized Columns” indexing
- 37. 37 FPGACPU User Application and Libraries CPU FPGA Interface Manager (FIM) Intel® Acceleration Engine with OPAE1 Technology Accelerator Function (Developer created or provided by Intel) UPI2/PCIe* HSSI3 Hypervisor & OS Optimized and simplified hardware and software APIs provided by IntelOPAE FPGA Acceleration Environment Common Developer Interface for Intel® FPGA Data Center Products Accelerator Function Interfaces
- 38. 38 Start developing for Intel® FPGAs with OPAE today: http://github.com/OPAE • Consistent API across product generations and platforms - Abstraction for hardware-specific FPGA resource details • Designed for minimal software overhead and latency - Lightweight user-space library (libfpga) • Open ecosystem for industry and developer community - License: FPGA API (BSD), FPGA driver (GPLv2) • FPGA driver being upstreamed into Linux* kernel • Supports both virtual machines and bare metal platforms • Faster development and debugging of Accelerator Functions with the included AFU Simulation Environment (ASE)** • Includes guides, command-line utilities and sample code FPGA Hardware + Interface Manager FPGA Driver (Physical Function – PF) FPGA API (C) (Enumeration, Management, Access) Applications, Frameworks, Intel® Acceleration Libraries Bare Metal Virtual Machine FPGA Driver (Virtual Function - VF) Hypervisor FPGA Driver (Common – AFU, Local Memory) OS Open Programmable Acceleration Engine (OPAE) Technology Simplified FPGA Programming Model for Application Developers
- 39. • Data analytics acceleration with no change to application required - Relational, NoSQL and Spark*/Hadoop* • Single workload with FPGA multifunction acceleration • Swarm64* accelerates big-and-fast data or traditional data warehousing • Potential to accelerate proxy access with partner rENIAC • Intel® FPGA card with framework for interfaces, end-to-end security and virtualization - Production in 2Q18 SUMMARY
Editor's Notes
- As businesses become more and more data intensive, the cost per transaction becomes an important metric. The combination of MariaDB and Intel® technologies is extremely powerful in this age of distributed computing.
- In this session, we will introduce the Intel/MariaDB collaboration team. Then we’ll discuss how the team is working to add shared, log-structured storage to support the persistence requirements of databases. Furthermore, you will learn how our cooperation supports the transformation of transaction performance and cost by optimizing MariaDB running on the combination of the Intel® Xeon® processor Scalable family and Intel® Optane™ SSDs. This platform includes the unique combination of 3D XPoint™ memory media with Intel’s advanced system memory controller, interface hardware, and software. We provide some early performance results based on an operational MariaDB/MyRocks implementation. We will also describe how Intel® FPGAs are accelerating database performance.
- Our work with MariaDB* is optimizing the combination of the following technologies: MariaDB and MyRocks* storage engine The Intel® Xeon® Scalable Platform using Intel® Optane™ SSDs This platform includes the unique combination of 3D XPoint™ memory media with Intel’s advanced system memory controller, interface hardware, and software.
- Intel is the ONLY company that powers every segment of the smart, connected world – from the cloud, to the network, to the device – and everything in between. We alone have the assets to power the next generation of technologies for the world. Our Strategy – A Virtuous Cycle We call it a virtuous cycle – the cloud and the data center, the Internet of Things, memory and FPGAs -- all bound together by connectivity and enhanced by the economics of Moore’s Law. Our strategy builds upon itself. These assets are all connected, and together, they fuel our business. With each new device that comes online and connects to the cloud – we have the tremendous opportunity to reinforce our company’s growth. The Cloud and Data Center The cloud is the most important trend shaping the future of the smart, connected world – and thus Intel’s future. Intel architecture defines the infrastructure of the cloud, and we will continue to drive more and more of the footprint of the data center globally. The cloud and analytics from the data center are the greatest value drivers in technology today. This digitization of everything will disrupt entire industries and open up new cycles of growth. At Intel, we’ll accelerate the power and value of this data and its analytics by continuing to innovate in high-performance computing, big data and machine learning capabilities. Internet of Things The Internet of Things encompasses all smart devices – every PC, device, sensor, console and any other edge device – that are connected to the cloud. The PC is foundational to our compute strategy and to our business. It’s an engine that creates critical shared IP that drives innovation across our entire product portfolio. Intel will continue to deliver an annual cadence of leadership performance and innovation in our PC and broader computing roadmap, with a focus on key growth opportunities in 2 in 1s, gaming and home gateways. We are imaging and inventing a PC that is truly immersive, assistive, with an incredibly natural interface tailored for your individual needs. Intel is also inventing a future that is more informed, collaborative, and meaningful given the widespread interconnectedness of everyday objects. As the Internet of Things evolves, we see three distinct phases emerging. Make everyday objects smart – this is well underway with everything from smart toothbrushes to smart car seats now available. Connect the unconnected, with new devices connecting to the cloud and enabling new revenue, services and savings. New devices like cars and watches are being designed with connectivity and intelligence built into the device. Deliver constant connectivity for devices that will need the intelligence to make real-time decisions based on their surroundings. At Intel, we will focus on autonomous vehicles, industrial and retail as our primary growth drivers of the Internet of Things. Memory and Programmable Solutions Memory and programmable solutions such as FPGAs will deliver entirely new classes of products for the data center and the Internet of Things. Intel® Rack Scale Design, 3D XPoint™ memory, FPGAs and silicon photonics are technologies that have been in development for several years at Intel and that we will bring to production soon. FPGAs are now at the heart of a diverse compute model that spans the Internet of Things (IoT), communications infrastructure, to the cloud and data center, and automotive markets. Intel FPGAs are used by over 10,000 customers in applications ranging from machine learning, to Advanced Driver Assist Systems (ADAS), from the core to edge, and in factory automation and smart energy systems among many others. Intel is driving innovations – using our deep understanding of materials science as well as computer architecture – to grow memory and storage at a much faster rate to virtually eliminate latency for data from storage. Bigger memory and faster storage benefits many new devices by enabling more immersive experiences with natural interaction and also provides significant value to the cloud by allowing businesses to run more efficiently. The availability of faster storage and bigger memory also unlocks more value in the cloud as we learn to automate and efficiently analyze increasing quantities of data. Connectivity Threading this virtuous cycle together is connectivity – the fact that providing computing power to a device and connecting it to the cloud makes it more valuable. In the future, we will add more than 50 billion smart and connected devices, machines, autonomous vehicles, buildings and cities. These devices will be always on and connected — with a demand for lower-latency where dynamic, split-second analytics, decisions and action on the data is required. At Intel, we recognize that 5G is more than an evolutionary step forward for our industry. As the world moves to 5G, Intel will lead because of our technological strength to deliver end-to-end 5G systems. This is why Intel is focusing on three key areas: industry partnerships, end-to-end 5G-related hardware and software development, and supporting 5G standards-setting. Intel will continue to develop technologies such as Mobile Edge Computing, millimeter wave and NarrowBand IOT (NB-IOT). These are all important steps to bring connectivity to a variety of new IoT devices globally. We’ll also continue to forge industry partnerships to develop and transform network infrastructure technologies. Moore’s Law Moore’s Law will continue to progress and Intel will confidently continue to harness its value. Intel’s leadership in Moore’s Law has driven the products delivering massive computing power growth and increasingly better economics and pricing. As we progress from 14 nanometer technology to 10 nanometer and plan for 7 nanometer and 5 nanometer and even beyond, our plans are proof that Moore’s Law is alive and well. Amazing Experiences Understanding this virtuous cycle, imagine a world where everything is smart and connected. Where your wedding ring monitors your blood sugar, your coffee pot automatically orders a refill of your favorite bean, or where your work station eliminates distractions to keep you productive - soon everything will seamlessly improve and support day-to-day tasks, both simple and complex. Where smart devices are embedded in all types of sports equipment, from jerseys, to balls, to protective gear. And these devices are generating vast amounts of data that are revolutionizing how players train, coaches teach, and scouts evaluate talent. The technologies we invent will also radically improve player safety, and transform the viewing experience. Audiences will be more engaged than ever, tracking the precise activity of their favorite athletes, from warm ups, to celebrating the victory. Where autonomous vehicles will soon be part of an integrated transportation network where cars will communicate with each other about traffic patterns, collisions, or even let you know when walking or taking mass transit is faster. Cars are becoming intelligent in a way that will save time and save lives. Intel transforms the technology we invent into amazing experiences.
- Current major sites where we are doing MariaDB related activities: Steve - HammerDB tool + remote testing Mikhail - running the lab and local testing
- Intel and MariaDB participate in several co-marketing collaborations, such as the Solution Brief, “Intel® Xeon® Platinum processor Accelerates MariaDB Server* and MyRocks*,” that you received in your Welcome Packet.
- Key Message: Introducing Xeon Scalable Platform Storyline: As I mentioned at the beginning, I’m excited to introduce the latest addition to our data center portfolio, the Intel Xeon Scalable Platform. This platform represents a new pinnacle in Intel’s 20+ year history of data center innovation. The Intel Xeon Scalable Processor was architected to deliver strong workload driven performance. Performance: We are excited to be delivering a consistent outstanding performance boost of 1.6X average across a wide range of real-world applications gen over gen. Compute, Storage, Network Optimized: It’s been a multi-year journey but finally, in this generation, we’ve implemented more features and capabilities specifically for storage and network than ever before. This enables us to win more workloads, more customers and enable new usage models or business opportunities. Security: We’re also lowering the performance overhead for security data; making data encryption / decryption without a performance penalty a reality. This allows data center operators to deploy security pervasively with minimal impact to service delivery. Simple and Easy to Deploy: In the last 5 years, we’ve done a lot of work with the ecosystem to ensure that all the capabilities that went into workload optimization are industry-ready for fast adoption. Also, for the first time, we are releasing a platform that is capable of 2, 4, & 8S+, all at the same time. This will not only accelerate the availability of larger systems in the industry but now, some of the mission critical features (previously available on higher-end system only) can now be supported in 2S mainstream systems.
- Efficient utilization of new enhanced processor cores and cache Ongoing opportunity for optimization for nonvolatile memory
- Software TCO 3 years Oracle Xeon E5-2697v3 14 Haswell Q3,14 - 28 cores $665,000 + 438900 22% support for 3 yrs = $1,103,900 Xeon E5-2699v4 22 Broadwell Q1,16 - 44 $1,045,000 + 689700 22% support for 3 years = $1,734,700 Xeon Platinum 8180 28 Skylake Q3,17 - 56 $1,330,000 + 877800 22% support for 3 years = $2,207,800 MariaDB Xeon E5-2697v3 14 Haswell Q3,14 - 28 cores $15000 Xeon E5-2699v4 22 Broadwell Q1,16 - 44 cores $15000 Xeon Platinum 8180 28 Skylake Q3,17 - 56 cores $15000 Moores Law MariaDB (Hammerdb oltp) Xeon E5-2697v3 387861 NOPM - $0.038 Cost per transaction Xeon E5-2699v4 490594 NOPM - $0.030 Cost per transaction Xeon Platinum 8180 659185 NOPM $0.022 Cost per transaction Performance v3 to v4 - 1.26X v4 to Platinum - 1.34X v3 to Platinum - 1.7X Oracle (Hammerdb oltp) Xeon E5-2697v3 918060 NOPM license $1.20 cost per transaction Xeon E5-2699v4 1404780 NOPM license $1.23 cost per transaction Xeon Platinum 810 2393498 NOPM 0.27X $0.92 cost per transaction Performance v3 to v4 - 1.53X v4 to Platinum - 1.7X Cost difference v3 32X v4 41X Platinum 42X
- Key Message: A glimpse inside the Xeon Scalable Platform Storyline: What’s inside this platform? It represents the best combination of leadership capabilities built upon Intel’s >20 year innovations across CPU and chipset and more. At the top-left, you see some of the technologies we will be offering in this generation as integrated options: Omnipath fabric, Ethernet - we’ll have Quick Assist integrated. Quick assist was born of network WL for compression and crypto acceleration. Over time, it has grown in use case where CSPs and Financial Services want to use it for their security. Another example is AVX-512. We’ve had it in other products and we are bringing it to Xeon in this generation. Hugely popular in HPC for acceleration and with FSIs for High Frequency Trading but AVX-512 benefits WLs across compute, storage, and network – we’ll show you more an real-world performance with AVX-512 optimizations in a few slides. These integrated options are for customers looking to reduce power or save space inside the system. Our platform supports our latest and greatest SSDs including the Optane SSDs as well as our complementary products like discrete Intel FPGAs. We have invested multiple years with the ecosystem on SW enablement to take advantage of the new core features and capabilities. Along with specialized workload frameworks like Caffe for artificial intelligence and development tools like Data Plane Development Kit for networking, we can accelerate the adoption of these new datacenter technologies and create new business opportunities. This platform delivers on the value vectors that matter to the customer to fully embrace the new market opportunities, accelerate growth and remain competitive: performance, security, and agility… Digging in a bit deeper… This platform provides additional choice of capability offerings and configurations. For example, for networking the Intel® Ethernet 700 series provides via an add-in card with support up to 40GbE for high speed data communications, supports enhanced Data Plane Development Kit (DPDK), which is a set of libraries and drivers to accelerate Network Functions Virtualization (NFV), and achieve higher performance for network packet processing workloads. This level of Ethernet capability is now offered via integration at the chip level & since DPDK is a core-intensive application, the better per core performance offered with the new Intel Xeon Processor scalable family over previous generation products enables better DPDK performance What about integrated accelerators like Intel® Advanced Vector Extensions 512 or Intel® QuickAssist? With new Xeon® processor families further extending the breadth of vector processing capability across both floating-point and integer data domains, and also adding Fused Multiply Add (FMA) support and scalability. Intel® AVX-512 can double the FLOPS/clock vs. AVX2. Intel QAT provides hardware acceleration for compute-intensive workloads (cryptography and data compression); To extend the platform value across the system we have new Intel® Data Center SSDs, which can be used across a continuum of data tiering needs including Storage, Caching and Memory. The new Intel® Optane™ SSDs are amazingly fast and the Intel® 3D NAND SSDs provide a terrific balance of speed and storage capacity. Both can benefit from a new platform capability we call Intel® Intel® Volume Management Device (Intel® VMD), which is designed to deliver seamless management of PCIe*-based (NVMe*) solid state drives, including enabling “hot plug” capability that minimize service interruptions during drive swaps. For specialized workload optimizations & evolving workloads like Artificial Intelligence, enterprises can benefit by combining the new Intel Xeon Scalable processors with other products from Intel’s portfolio, ranging from FPGAs and the Intel® Xeon® Phi processor to the Intel Nervana offerings. Snap*: The Open Telemetry Framework (Snap*): Data is everywhere around us. Telemetry, said simply, is system information: anything and everything we can gather related to the state of a process, piece of hardware, OS, virtualization layer and cloud software. This information is diverse in form and its effectiveness in driving decisions. Individual data points don’t tell much of a story. In aggregate, it can be truly insightful. Enter Snap, the open telemetry framework that provides greater exposure to system data by standardizing telemetry behind a single API. Storage Performance Development Kit (SPDK): To help storage OEMs and ISVs integrate this hardware, Intel has created a set of drivers and a complete, end-to-end reference storage architecture called the Storage Performance Development Kit (SPDK). The goal of SPDK is to highlight the outstanding efficiency and performance enabled by using Intel’s networking, processing, and storage technologies together. By running software designed from the silicon up, SPDK has demonstrated that millions of I/Os per second are easily attainable by using a few processor cores and a few NVMe drives for storage with no additional offload hardware.
- The P4800X is the World’s Most Responsive Data Center SSD. The Intel Optane SSD DC P4800X brings an industry-leading combination of low latency, outstanding QoS, and high endurance, delivering performance needed for both Memory and Storage workloads. In storage applications, the Optane SSD will help Break Storage and caching bottlenecks, enabling faster performance and increasing scalability. The Intel Optane SSD will also be used to replace or extend the DRAM memory pool, enabling more affordable deployments and significantly larger data sets. With these larger data sets closer to the Processor, we will see new insights and discoveries. As you all know, at Intel, we always ensure that we have data to back up any claims. Let’s dive into the elements and data which support this bold claim.
- Notes from DCG SDI Gold Deck. Storage is probably the biggest near-term need in Data Centers. The solutions are rapidly maturing, and the value proposition is strong. Reduction in the cost per Terabyte of storage by a factor of 3 to 15 X. (Swiftstack case study) The major changes as we move from traditional storage to SDI storage are: BEFORE: Each major system has dedicated storage resources, including disks, networks, back-up and archives. AFTER: Storage resources are pooled across systems, with shared infrastructure, reducing cost and wasted capacity. What had been fixed-function devices become virtual appliances running on standard Xeon servers. BEFORE: Each system kept stored data in isolated vaults, which made cross-system or cross-enterprise analysis very slow and difficult. AFTER: Unified resource pools with a single namespace managed by a master orchestrator enables greater visibility into stored data, and business insights yielded from simpler analysis. BEFORE: Capacity was delivered through expensive scale-up systems, that often locked customers into specific vendors. AFTER: Scale-out systems allow customers to easily add capacity to their storage pools. And since they are based on standard Intel Xeon platforms, customers have more choices and lower cost. All this, means customers’ storage systems become much more nimble and easy to configure, reduce total cost, and opens up their stored data to more analysis and insight.
- https://www.slideshare.net/Sheeri/scale-db-preso-for-boston-my-sql-meetup-92009
- Dong et al, “Optimizing Space Amplification in RocksDB,” 2017 http://cidrdb.org/cidr2017/papers/p82-dong-cidr17.pdf MyRocks - a RocksDB storage engine with MySQL http://myrocks.io/ Matsunobu, "MyRocks: A space- and write-optimized MySQL database," 2016 https://code.facebook.com/posts/190251048047090/myrocks-a-space-and-write-optimized-mysql-database/
- Cloud-centric, shared storage architecture
- Three imperatives enable the process of data center modernization. Enable visibility, virtualization and performance management through cloud. Engaging cloud with visibility across all your workloads allows you to optimize your operations and flexibly respond to future needs while guaranteeing SLAs. Only about half of IT infrastructures are virtualized, leaving plenty of opportunities to consolidate and improve operations. Migrate to a services-oriented business through software-defined infrastructure (SDI). Software-defined infrastructure is the data center approach that will enable fast services delivery by quickly configuring resources to meet the needs of new services. A self-service portal allows users to choose their own services, so they don’t have to wait for technologies they can leverage to innovate with. This will drive down time to market, costs and labor. Create greater value in IT by driving business innovation with analytics. With the growing data that companies acquire, IT is in a position to leverage that data to gain insight that drives innovation.
- A Reconfigurable Fabric for Accelerating Large-Scale Datacenter Services Microsoft's Production Configurable Cloud Accelerating Persistent Neural Networks at Datacenter Scale Average 7.72 usec to L1 switch and average 18.71 usec latency to L2 switch Roughly half height half length PCIe expansion card ; 29.2W maximum power consumption
- Relational: 2X+ TPC-DS or TPC-H w/Swarm64 NoSQL: 4X Cassandra w/rENIAC (80/20 R/W) Hadoop/Spark: 3X Terasort w/A3Cube (HDD) As an example: Relational: 2X+ TPC-DS or TPC-H w/Swarm64 • Microsoft: A Reconfigurable Fabric for Accelerating Large-Scale Datacenter Services • Microsoft's Production Configurable Cloud • Accelerating Persistent Neural Networks at Data Center Scale • Average 7.72 usec to L1 switch and average 18.71 usec latency to L2 switch • Roughly half height half length PCIe expansion card ; 29.2W maximum power consumption
- Banking/Finance, Business Intelligence, Government, Healthcare, Retail
- At the heart of the Acceleration Stack is the Acceleration Environment for Intel Xeon CPU with FPGAs. This Acceleration Environment has a software component, the Intel Acceleration Engine, and an FPGA component, the FPGA Interface Manager. Together, these provide the common developer interface with simplified and optimized APIs and interfaces from Intel. Developer can build their own acceleration applications and Accelerator Functions directly on top of the Acceleration Environment and reuse that code on different Intel FPGA form factors in the data center. Or they can take advantage of the Acceleration Libraries for FPGAs to fast-track their performance and time to market.
- Intel® Programmable Acceleration card with Intel Arria® 10 FPGA Sampling now, production 2Q18; offering end-to-end virtualization and security