peak performance
play

Peak Performance Remote Memory Revisited Hannes Mhleisen, Romulo - PowerPoint PPT Presentation

Peak Performance Remote Memory Revisited Hannes Mhleisen, Romulo Goncalves and Martin Kersten Database Scalability Scale-Up Scale-Out (Big Iron) (Many Boxes) Cheating Full Virtualization Storage Clusters Remote Memory 2 Why


  1. Peak Performance Remote Memory Revisited Hannes Mühleisen, Romulo Goncalves and Martin Kersten

  2. Database Scalability Scale-Up Scale-Out (Big Iron) (Many Boxes) “Cheating” Full Virtualization Storage Clusters Remote Memory 2

  3. Why more memory? • Memory is a critical resource, especially in OLAP use cases • Hash tables, intermediate results, ... • OS overcommits, leads to thrashing 3

  4. 70 RAX XT ● 60 US$/GB 50 T1650 ● 40 T5600 ● ● T3600 ● RAX − SX4 ● ● 30 RAX − XS3 ● ● RAX − XS4 ● 0 500 1000 1500 2000 Main Memory (GB) 4

  5. Remote memory then • Hack Kernel to page out to remote machines? [Tell et al. 2013] • Store swapfile to remote file system? • But DBs like to avoid swap anyway, so... • Store DB temporary files on remote system! 5

  6. New Toys 6

  7. The way it was RDMA CPU CPU Memory Memory Network Adapter Network Adapter Network Adapter Network Adapter Memory CPU Memory CPU Many-Copy Zero-Copy 7

  8. 8

  9. Experimental Setup • 14 Linux COTS Boxes • 16 GB RAM • InfiniBand QDR • 182 GB Memory total (and usable!) 9

  10. Throughput (GB/s) 0 1 2 3 HDD SSD Throughput iSCSI/ETH iSCSI/EoIB iSCSI/RDMA 10 NFS/ETH NFS/EoIB NFS/RDMA NFS/RDMA/RAID0 NFS/RDMA/RAID5 Read Write

  11. Latency (µs) 200 400 600 0 SSD RAM iSCSI/ETH Latency iSCSI/EoIB iSCSI/RDMA 11 NFS/ETH NFS/EoIB NFS/RDMA NFS/RDMA/RAID0 NFS/RDMA/RAID5

  12. OLAP DB (TPC-H) • TPC-H: benchmark for relational databases focused on analytics (OLAP) • Queries tend to have large intermediate results (SF=100): Query Read (GB) Write (GB) 1 14 50 18 5 28 21 7 9 3 6 6 13 2 7 12

  13. Example: Query 18 2.0 Traffic (GB/s) 1.5 Direction Read 1.0 Write 0.5 0.0 0 30 60 90 Time (s) 13

  14. TPC-H Experiment • Single node runs MonetDB with TPC-H database (SF=100) • Gets remote memory from the 14 memory providers • DB temporary partition resides either on disk or in remote memory • Hot runs, 5 repetitions per query and setup 14

  15. Average Execution Time (s) 1000 10 TPC-H 100 Results q01 q02 q03 q04 q05 q06 q07 q08 q09 q10 Query q11 15 q12 q13 q14 q15 q16 q17 q18 q19 q20 q21 q22 Experiment RRAM HDD

  16. Summary • Remote Memory is interesting (...) • Lightweight technique • RDMA allows for remote memory to make sense from a technical perspective • OLAP database scenarios can benefit from this • Open issue: Hardware pricing/TCO 16

  17. Thank You! Questions? http://is.gd/remotemem

Recommend


More recommend