Understanding DPDK
•Download as PPTX, PDF•
188 likes•103,763 views
1. DPDK achieves high throughput packet processing on commodity hardware by reducing kernel overhead through techniques like polling, huge pages, and userspace drivers. 2. In Linux, packet processing involves expensive operations like system calls, interrupts, and data copying between kernel and userspace. DPDK avoids these by doing all packet processing in userspace. 3. DPDK uses techniques like isolating cores for packet I/O threads, lockless ring buffers, and NUMA awareness to further optimize performance. It can achieve throughput of over 14 million packets per second on 10GbE interfaces.
More Related Content
What's hot
What's hot (20)
Viewers also liked
Viewers also liked (20)
Similar to Understanding DPDK
Similar to Understanding DPDK (20)
More from Denys Haryachyy
More from Denys Haryachyy (6)
Recently uploaded
Recently uploaded (20)
Understanding DPDK
- 1. Understanding DPDK Description of techniques used to achieve high throughput on a commodity hardware
- 2. How fast SW has to work? 14.88 millions of 64 byte packets per second on 10G interface 1.8 GHz -> 1 cycle = 0,55 ns 1 packet -> 67.2 ns = 120 clock cycles IFG Pream ble DST MAC SRC MAC SRC MAC Type Payload CRC 84 Bytes 412 8 60
- 3. Comparative speed values CPU to memory speed = 6-8 GBytes/s PCI-Express x16 speed = 5 GBytes/s Access to RAM = 200 ns Access to L3 cache = 4 ns Context switch ~= 1000 ns (3.2 GHz)
- 4. Packet processing in Linux User space Kernel space NIC App Driver RX/TX queues Socket Ring buffers
- 5. Linux kernel overhead System calls Context switching on blocking I/O Data copying from kernel to user space Interrupt handling in kernel
- 6. Expense of sendto Function Activity Time (ns) sendto system call 96 sosend_dgram lock sock_buff, alloc mbuf, copy in 137 udp_output UDP header setup 57 ip_output route lookup, ip header setup 198 ether_otput MAC lookup, MAC header setup 162 ixgbe_xmit device programming 220 Total 950
- 7. Packet processing with DPDK User space Kernel space NIC App DPDK Ring buffers UIO driver RX/TX queues
- 8. Kernel space Updating a register in Linux User space HW ioctl() Register syscall VFS copy_from_user() iowrite()
- 9. Updating a register with DPDK User space HW assign Register
- 10. What is used inside DPDK? Processor affinity (separate cores) Huge pages (no swap, TLB) UIO (no copying from kernel) Polling (no interrupts overhead) Lockless synchronization (avoid waiting) Batch packets handling SSE, NUMA awareness
- 11. Linux default scheduling Core 0 Core 1 Core 2 Core 3 t1 t4t3t2
- 12. How to isolate a core for a process To diagnose use top “top” , press “f” , press “j” Before boot use isolcpus “isolcpus=2,4,6” After boot - use cpuset “cset shield -c 1-3”, “cset shield -k on”
- 13. Core 2Core 1 Run-to-completion model RX/TX thread RX/TX thread Port 1 Port 2
- 14. Core 2Core 1 Pipeline model RX thread TX thread Port 1 Port 2 Ring
- 15. Page tables tree Linux paging model cr3 Page Page Global Directory Page Table Page Middle Directory
- 17. TLB characteristics $ cpuid | grep -i tlb size: 12–4,096 entries hit time: 0.5–1 clock cycle miss penalty: 10–100 clock cycles miss rate: 0.01–1% It is very expensive resource!
- 18. Solution - Hugepages Benefit: optimized TLB usage, no swap Hugepage size = 2M Usage: mount hugetlbfs /mnt/huge mmap Library - libhugetlbfs
- 19. Lockless ring design Writer can preempt writer and reader Reader can not preempt writer Reader and writer can work simultaneously on different cores Barrier CAS operation Bulk queue/dequeue
- 20. Lockless ring (Single Producer) 1 cons_head cons_tail prod_head prod_tail prod_next 2 cons_head cons_tail prod_head prod_next prod_tail 3 cons_head cons_tail prod_head prod_tail
- 21. Lockless ring (Single Consumer) 1 cons_head cons_tail prod_head prod_tail cons_next 2 cons_tail prod_head prod_tail cons_next cons_head 3 cons_head cons_tail prod_head prod_tail
- 22. Lockless ring (Multiple Producers) 1 cons_head cons_tail prod_head prod_tail prod_next1 prod_next2 3 cons_head cons_tail prod_head 2 cons_head cons_tail prod_head prod_next2 prod_tail prod_next1 4 cons_head cons_tail 5 cons_head cons_tail prod_head prod_tail prod_tail prod_head prod_tail prod_next1 prod_next2 prod_next1 prod_next2
- 23. Kernel space network driver App IP stack Driver NIC Data Desc Config Data User space Kernel space Interrupts
- 24. UIO “The most important devices can’t be handled in user space, including, but not limited to, network interfaces and block devices.” - LDD3
- 25. UIO User space Kernel space Interfacesysfs /dev/uioX App US driver epoll() mmap() UIO framework driver
- 26. NIC User space Access to device from user space BAR0 (Mem) BAR1 BAR2 (IO) BAR5 BAR4 BAR3 Vendor Id Device Id Command Revision Id Status ... Configuration registers I/O and memory regions /sys/class/uio/uioX/maps/mapX /sys/class/uio/uioX/portio/portX /dev/uioX -> mmap (offset) /sys/bus/pci/devices
- 27. Host memory NIC memory DMA RX Update RDT DMA descriptor(s) RX queue RX FIFO DMA packet Descriptor ringMemory DMA descriptors
- 28. Host memory NIC memory DMA TX Update TDT DMA descriptor(s) TX queue TX FIFO DMA packet Descriptor ringMemory DMA descriptors
- 29. Receive from SW side DD DD DDDD RDT DD mbuf1 addr DD mbuf2 addr RDT RDH = 1 RDT = 5 RDBA = 0 RDLEN = 6 mbuf1 RDH RDH mbuf2
- 30. Transmit from SW side DD DD DDDD TDT DD mbuf1 addr DD mbuf2 addr TDT TDH = 1 TDT = 5 TDBA = 0 TDLEN = 6 mbuf1 TDH TDH mbuf2
- 31. NUMA CPU 0 Cores Memory controller I/O controller Memory PCI-E PCI-E CPU 1 Cores Memory controller I/O controller Memory PCI-E PCI-E QPI Socket 0 Socket 1
- 32. RSS (Receive Side Scaling) Hash function Queue 0 CPU N ... Queue N Incoming traffic Indirection table
- 33. Flow director Queue 0 CPU N ... Queue N Incoming traffic Filter table Hash function Outgoing traffic Drop Route
- 34. Virtualization - SR-IOV NIC VMM VM1 VF driver VM2 VF driver PF driver VF Virtual bridge VF PF
- 35. NIC Slow path using bifurcated driver Kernel DPDK VF Virtual bridge PF Filter table
- 36. Slow path using TAP User space Kernel space NIC App DPDK Ring buffers TAP device RX/TX queues TCP/IP stack
- 37. Slow path using KNI User space Kernel space NIC App DPDK Ring buffers KNI device RX/TX queues TCP/IP stack
- 38. x86 HW Application 1 - Traffic generator User space Streams generator DUT Traffic analyzer
- 39. x86 HW Application 2 - Router Kernel User space Routing table Routing table cacheDUT1 DUT2
- 40. x86 HW Application 3 - Middlebox User space DPIDUT1 DUT2
- 41. References Device Drivers in User Space Userspace I/O drivers in a realtime context The Userspace I/O HOWTO The anatomy of a PCI/PCI Express kernel driver From Intel® Data Plane Development Kit to Wind River Network Acceleration Platform DPDK Design Tips (Part 1 - RSS) Getting the Best of Both Worlds with Queue Splitting (Bifurcated Driver) Design considerations for efficient network applications with Intel® multi-core processor-based systems on Linux Introduction to Intel Ethernet Flow Director
- 42. My blog Learning Network Programming