“We want to use machine learning and advanced analytics to gain insights and do more with the flood of useful data at our fingertips, but our infrastructure can't keep up.”
Sound familiar? I hear this lament every week as I talk to enterprise customers working with a wide range of workloads, from financial trading to healthcare to e-commerce to AI modeling.
Database and IT admins recognize the enormous opportunities available to their businesses from big data, but they feel hampered by the limitations of traditional data storage. SSDs—even faster 3D NAND SSDs—aren’t sufficient for real-time processing of hot data. Adding DRAM is an option, but it's costly and limited in capacity.
Now, NetApp Memory Accelerated Data (MAX Data) offers businesses an alternative. NetApp MAX Data is built with Intel® Optane™ persistent memory (PMem), a revolutionary memory technology in a DIMM form factor that brings hot data closer to the CPU for faster performance with low latency. Intel Optane PMem offers two key benefits:
- High-capacity memory, with support for up to 512 GB per DIMM at a lower cost than DRAM
- Full data persistence, unlike volatile DRAM
Intel Optane PMem can be used in either of two modes. The first, Memory Mode, offers traditional, volatile memory that provides high-capacity performance, without the need to rewrite any application code. The second, App Direct Mode, adds data persistence and can be used with applications that have been rewritten to support byte-addressable memory, instead of traditional, block-addressable storage.
Several popular software vendors, including SAP, Oracle, Microsoft, and others, already offer applications designed to take advantage of Intel Optane PMem in App Direct Mode. But what if you have a home-grown or legacy application, or your preferred software doesn't yet support App Direct Mode?
That's where NetApp MAX Data comes in. MAX Data is software you can seamlessly install directly on your application server to make all the features of Intel Optane PMem—including data persistence in App Direct Mode—available to all the applications on that server. There's no need to rewrite any code.
MAX Data is not a caching system that makes a copy of hot data in memory for fast read access. Instead, it’s a tiering system that makes the memory tier the primary location for hot data for fast read and write operations. As the memory tier fills, cooler data is automatically and transparently moved (not copied) to the warm data-storage tier.
MAX Data Offers up to 3X Higher Performance in Testing
We wanted to see for ourselves how much difference NetApp MAX Data with Intel Optane PMem would make in real-world use cases, so we put it to the test. We ran a series of benchmark test scenarios of Oracle Database on bare-metal and virtualized environments.
We used HammerDB on a bare-metal configuration with Oracle Database and compared throughput for systems with and without MAX Data. The results showed up to 1.9 times better transactional throughput after adding MAX Data to the bare-metal system.[i] When we ran the same tests on a virtualized environment (the database instances were in VMs), the difference was even greater, peaking at up to 3.16 times more transactions per minute.[ii]
We worked with NetApp engineers to run several additional tests on a wide range of SQL and NoSQL database applications. In every scenario, the systems configured with MAX Data and Intel Optane PMem showed higher performance than the baseline systems. Check out the white paper for details and full results.
If you’re ready to transform your data center to support modern, data-driven workloads like AI, NetApp MAX Data can help you get there today by offering the full benefits of Intel Optane technology without the need to rewrite applications.
- Get the full story in our white paper.
- Check out the webcast I recently presented with our partners at NetApp.
[ii] Oracle Database on virtualized environment with HammerDB benchmarking software: Based on testing by Intel on November 4, 2019. XFS configuration: Intel Server Board S2600WF, Intel Xeon Gold 6252 processor (24 cores/socket, 2 sockets, 2 threads per core), 12 x 32 (384 GB total) 2,667 MHz DDR4 ECC DRAM, network: dual-port Mellanox ConnectX-4 Lx (25 gigabit Ethernet [GbE], bonded), storage: operating system: 1 x 1.6 TB Intel SSD DC S3610, database: 2 x 8 TB Intel SSD DC P4510 PCIe, RAID 0 (data) + 1 x 8 TB Intel SSD DC P4510 (redo); operating system: Fedora 29, BIOS: WW26 (SE5C620.86B.02.01.0008.031920191559), microcode: 05000017, running Oracle Database 19c Enterprise Edition release 220.127.116.11.0—production version 18.104.22.168.0, HammerDB 3.2, database size = 2 x 1 TB on XFS file system; virtualization: QEMU/KVM 4.0.94 (v4.1.0-rc4); 2 guest VMs with 160 GB vDRAM, 48 vCPUs each, CentOS 7.6, 4 TB storage for database. Database cache size = 100 GB. NetApp MAX Data configuration: Intel Server Board S2600WF, Intel Xeon Gold 6252 processor (24 cores/socket, 2 sockets, 2 threads per core), 12 x 32 (384 GB total) 2,667 MHz DDR4 dual-rank ECC DRAM, 8 x 128 GB (1 TB total) Intel Optane PMem in a 2-2-1 configuration, network: dual-port Mellanox ConnectX-4 Lx (25 GbE, bonded), storage: operating system: 1 x 1.6 TB Intel SSD DC S3610, database: 2 x 8 TB Intel SSD DC P4510 PCIe (data) + 1 x 8 TB Intel SSD DC P4510 (redo); operating system: Fedora 29, BIOS: WW26 (SE5C620.86B.02.01.0008.031920191559), Intel Optane PMem firmware: 01.02.00.5410, microcode: 05000017, running Oracle Database 19c Enterprise Edition release 22.214.171.124.0—production version 126.96.36.199.0, HammerDB 3.2, database size = 2 x 1 TB on NetApp MAX FS 1.5 file system; virtualization: QEMU/KVM 4.0.94 (v4.1.0-rc4); 2 guest VMs with 160 GB vDRAM, 48 vCPUs each, 1 TB Intel Optane persistent memory, CentOS 7.6, 4 TB storage for database. Database cache size = 100 GB.