- people - netdevconf.info

0x17: Talk, Nanotube: Turning XDP Code into Custom Hardware Pipelines for FPGA SmartNICs
by Jamal Hadi Salim 07 Oct '23

07 Oct '23

Many modern NICs have built-in packet processing capabilities that can effectively perform the type of use case that XDP is often used for. Stephan Diestelhorst, Neil Turton, Kieran Mansley, Kimon Karras, Rip Sohan , Thomas Calvert and Steven Pope say that being able to seamlessly offload this XDP code to the NIC – under the control of the Linux networking subsystem and tooling – would increase server efficiency and performance without affecting the user experience. Stephan Diestelhorst et al will discuss Nanotube: an open-source compiler that consumes standard XDP code and produces a tailor-made, application-specific dataflow pipeline suitable for execution on FPGA NIC platforms. They have demonstrated generated Katran on AMD/Xilinx FPGA-based NICs to offload with throughput at line rate (100 Gbps) on AMD NICs! In this talk they will describe (i) the structure of the compiler; (ii) performance of typical workloads; and (iii) how map accesses by host software are supported using specially annotated packets over the normal host data path. And last but not least they will discuss requirements to allow better integration and support by the Linux kernel, in particular the validator, map handling, and vendor-neutral support requirements. cheers, jamal

1 0

0x17: Talk, P4 Compatible Control Plane Driver for Linux Kernel
by Jamal Hadi Salim 07 Oct '23

07 Oct '23

Our friends at Intel and AMD will share their thoughts on how to integrate P4TC[1] into their drivers. In this talk Anjali Singhai, Neha Singh, James Choi, Namrata Limaye, Sridhar Samudrala, Mahesh Shirshyad and Vipin Jain will discuss enhancements to existing TC offload interfaces to accommodate multiple P4 pipeline and runtime control. They will elaborate on the association of the s/w and h/w datapaths, hardware P4 dataplane loading and initialization as well as runtime control (using the classical skip_sw/hw flagging). These details will include describing the kernel infrastructures and any extension to current TC offload. Anjali et all will discuss potential enable shimming of legacy kernel offloads such as tc-flower on to P4TC interfaces. Additionally, they will talk about interfaces to support legacy hooks (switchdev ops, FDB ops) and coexistence of these with P4TC compatible driver. They will make the infrastructure code at the conference. cheers, jamal [1] https://netdevconf.info/0x17/sessions/talk/integrating-ebpf-into-the-p4tc-d…

1 0

0x17: Talk, SO_TIMESTAMPING: powering fleetwide RPC monitoring
by Jamal Hadi Salim 07 Oct '23

07 Oct '23

Willem de Bruijn says timestamping, via SO_TIMESTAMPING, is key to debugging network stack latency. Instead of gut-feel finger pointing between network and kernel tribes we can just get down to the facts of where the latency really is. SO_TIMESTAMPING can isolate transmission, reception and even scheduling sources. Capturing connection state along with timestamps further enables root cause discovery, such as TCP receive window size. Capturing timestamps at more points, such as traffic shaping and NIC hardware, expands visibility to tough issues like incast. SO_TIMESTAMPING has seen iterative development to enable fleetwide RPC monitoring at Google. They presented details of this infra called "Fathom" at SIGCOMM 2023[1]. In this talk Willem will start us beyond where that SIGCOMM paper ends. He will take a deep dive on the Linux kernel infrastructure that makes fleetwide continuous latency analysis and attribution possible. API extensions include covering TCP bytestreams, capturing transport protocol state along with events (OPT_STAT), and supporting selective sampling (OPT_CMSG). The talk reviews the core SO_TIMESTAMPING API, discusses non-obvious extensions (MSG_EOR, SO_RCVLOWAT), summarizes gotchas from the field (OPT_ID_TCP), and explains how all this combines to enable robust continuous RPC monitoring. It touches on clock synchronization and precision. Finally, it compares this UAPI to dynamic tracing with uprobes, kprobes, tracepoints and BPF. Should be fun, cant wait to attend this talk! [1] search for "fathom" at https://conferences.sigcomm.org/sigcomm/2023/program.html cheers, jamal

1 0

Assessing the impact of Linux networking on CPU consumption
by Jamal Hadi Salim 07 Oct '23

07 Oct '23

You may be a guru at reading perf output and then with a handwave, swish, you can whip your bash script into a frenzy to create flame graphs... but you would still need to dig deeper and isolate where your bottlenecks in the network are. Davide Miola, Fulvio Risso and Federico Parola took a different approach by creating a tool using eBPF to instrument strategic functions in the network RX and TX entry-points. They crafted a novel tool that they call "Netto"[1] which gives you a focussed view of only the network infra's contribution. Netto can give you answers to coarse grained details of "how much is the rx softirq chewing vs RX syscall" all the way to more fine-grained rx softirq details like the driver cost, GRO, NF, etc. And then do all this in real time with low cpu overhead... In this talk, Davide et al will go into details of the design and functionality of Netto as well as describe the different kernel hook points. They will then illustrate how one would be able to use this tool to get insights into the network infra's overheads for a specific workload. [1]https://github.com/miolad/netto cheers, jamal

1 0

0x17: Talk, Scripting the Linux Routing Table with Lua
by Jamal Hadi Salim 06 Oct '23

06 Oct '23

You can run Lua scripts inside the kernel using the Lunatik project! Lunatik has been previously used for scripting Netfilter and XDP subsystems, allowing users to create complex network filters using Lua. For instance, it has been used for filtering L7 traffic such as HTTP and HTTPS inside the kernel. In this talk, Lourival Vieira Neto, Marcel Moura, Ana Lúcia de Moura and Roberto Ierusalimschy present the evolution of this framework and illustrate a sample use case implementing a kernel-level adaptive routing service. This service allows the system to monitor the reliability of a network route and adjust its routing table dynamically in case of failure. For instance, a home router might use such service to fall back to the cellular network using a mobile phone as a gateway. Lourival et al also spice this service up by implementing a protocol that allows nodes to advertise themselves as alternative gateways and for the router to notify nodes when elected as the new gateway. Lourival et al will describe the kernel details and associate infrastructure interaction ( netdevice notification chain, kthread, socket, RCU and FIB, etc). cheers, jamal

1 0

0x17: Talk, eBPF Qdisc: a generic building block for traffic control
by Jamal Hadi Salim 06 Oct '23

06 Oct '23

Adding a new packet scheduling algorithm in the kernel requires implementing a Qdisc kernel module - which could be cumbersome. Writing a custom scheduling algorithm which may not make sense for the rest of the community is frowned upon. On top of Qdiscs, different classifiers and rules can be applied to packets as well, which further complicates the combination of choices. Amery Hung and Cong Wang introduce eBPF Qdisc, a fully programmable Qdisc with eBPF. Amery and Cong believe that the approach they are taking will provide developers not only a more flexible and safer way to implement packet schedulers, but also orchestrating packets in coordination with other networking components. In this talk, they first show that eBPF Qdisc is capable of realizing a sophisticated Qdisc by implementing fair queueing (fq) Qdisc. Then, they evaluate the performance of the fq against the existing fq in the Linux kernel. Finally, they will show how eBPF Qdisc can help developers implement Qdisc's with diverse functionalities through two use cases: 1) we build a more robust earliest departure time (EDT) rate limiter by coordinating with a fq Qdisc through eBPF maps and 2) show eBPF Qdisc as a flexible way to create network emulators. cheers, jamal

1 0

0x17 Talk, Leveraging Homa: Enhancing Datacenter RPC Transport Protocols
by Jamal Hadi Salim 06 Oct '23

06 Oct '23

Hyper-scale data centers, such as Bytedance, have a strong need for optimal RPC performance. Homa is purpose-built for RPCs making it ideal for such environments. In this talk, Xiaochun Lu and Zijian Zhang do a comparative study of gRPC's performance on Homa versus over TCP. They observe clear performance advantages in favor of Homa, especially for RPC messages under 50KB, in terms of both latency and throughput. Like John's talk[1], they optimize longer RPC messages and improve bidirectional messaging. Xiaochun and Zijian will further provide performance insights, describe their mods to the Homa kernel code, and guide participants through the tangible benefits of integrating Homa into their RPC transport framework. [1]https://netdevconf.info/0x17/sessions/talk/kernel-managed-user-buffers-in-homa.html cheers, jamal

1 0

0x17: Talk, Kernel-Managed User Buffers in Homa
by Jamal Hadi Salim 06 Oct '23

06 Oct '23

Reminder: One more day to go for early bird registration.. Homa was first introduced to us in John Ousterhout's exciting 0x16 keynote[1]. John argued in the keynote that it was time to replace TCP with Homa to avoid what he labelled as the "datacentre tax". For the uninformed: Homa is message-oriented like UDP or SCTP as opposed to TCP being stream-oriented. There is no notion of connection in Homa, only request/response RPC messaging. A single Homa socket can send/receive RPCs to many peers. One advantage that TCP has by virtue of using the streaming approach is it can pipeline very large messages (think 500KB across MTU sized packets on the wire) - whereas a message oriented approach will have to copy bytes to/from user space and DMA-able packet buffers in parrallel. In his experiments with an mlx5 based 25G NIC, Dr. Ousterhout found that TCP infact outperformed Homa with large messages. At 500KB message size TCP was able to achieve 17-19 Gbps throughput, Homa was only able to hit 10-11Gbps. How dare TCP beat Homa in any metric! So John set out to fix this situation ;->. When he was done with the preliminary work Homa was able to hit the same rate as TCP (17-19Gbps). In this talk, he will go over the approach he took in the kernel code to allow the kernel to manage user mmaped buffers, changes made to both recvmsg() and sendmsg() to make this happen. 1. https://netdevconf.info/0x16/sessions/keynote/keynote-ousterhout.html cheers, jamal

1 0

0x17: Talk, Zero Copy Receive using io_uring
by Jamal Hadi Salim 05 Oct '23

05 Oct '23

The war on memory bottlenecks[1][2] continues with this talk from David Wei and Pavel Begunkov. Socket recv() data is first copied/DMAed into the kernel memory and then again into user space memory - adding pressure to overall memory bandwidth and of course comes with a CPU cost. While there are other approaches to avoid the second copy on recv() such as DPDK, RDMA etc, David and Pavel argue that all of them have downsides - ranging from being proprietary, custom patched drivers, and difficult to debug and worse requiring a rewrite of the applications... Our esteemed speakers instead opt for continuing to use the network stack as is and io_uring as the user facing API. So what do we need to get this working? As in [2] that the NIC supports header splitting and RSS flow steering. On incoming data the headers traverse the TCP stack while data is DMA'ed not to the kernel memory but into user space... David and Pavel will discuss the overall approach they took and describe in some detail the kernel infra as well as what uapi would look like. They will further review kernel-existing zero copy approaches and how they plan to coexist with them. Last but not least they will dig into limitations and challenges of zero copy receive and how they overcome them to facilitate a real deployment. [1] https://netdevconf.info/0x17/sessions/talk/congestion-control-architecture-… [2] https://netdevconf.info/0x17/sessions/talk/device-memory-tcp.html cheers, jamal Reminder: 2 more days to go for early bird registration

1 0

0x17: Talk, Device Memory TCP
by Jamal Hadi Salim 05 Oct '23

05 Oct '23

My apologies for putting some marketing at the top of this email: Reminder, Early bird ticketing is open for 2 more days; the sooner you register the better it is for our planning... And now to tell you about one of my favorite talks this time.. ML/AI continues to be an inspiration for new Linux kernel infrastructure and a theme at netdev conf 0x17! There are a couple of patterns in ML traffic that inspire this work from Mina Almasry, Willem de Bruijn, Eric Dumazet and Kaiyuan Zhang: 1) There is _a lot of freaking data_ being transferred across the network - and it continues to increase dramatically (ex: to accommodate such large data transfers NICs with 800 Gps ports are starting to appear) and 2) The vast majority of this data is typically from a device(eg storage) to another device (eg GPU). For context, assume a scenario where data transfers from SSDs on machineA cross the network to GPUs on machineB; and on machine B: In the rx direction incoming data into the NIC goes to host memory, then from host memory to the GPU. In the tx direction outgoing data goes from GPU to host then out to the NIC. This data movement approach, with multiple round trips across the memory and PCIe busses, is no longer tenable at such large amounts of data quantity. Current host hardware is just not equipped to deal with this. While such bulk transfers amount to low packets/sec rates, you are simply not going to be able to achieve wire speed because of the PCIe and memory bandwith abuse from the multiple round trips (see talk [1] which delves into host limitations). Almasry et al introduce Device Memory TCP to solve these challenges. They implement socket API to enable a user to send directly from device memory (GPU in machineB above example or SSD in machineA) across the network - as well as be able to place incoming data directly in the device memory (GPU in machineB example). How does this magic happen? Well, we need some support from the NIC to split the header on rx such that the TCP headers go to the host and the rest of the data goes directly to the device. The reverse is true for tx direction. Packet headers get to go to stack and get processed in the standard (TCP) code path. Memory bandwidth use is cut dramatically and because NIC to device transfers now happen at the lowest level of the PCIe tree hierarchy PCIe bandwidth use is also cut down dramatically. With devmem TCP, Almasry et al with a setup of 4 machines of GPU+100Gbps NICs - and data sent and received directly from/to device memory - were able to reach ~96.6% line rate speeds. In the talk they will go over details of the kernel/uapi changes to achieve devmem TCP. cheers, jamal [1] https://netdevconf.info/0x17/sessions/talk/congestion-control-architecture-…

1 0

0x17: Talk, Lightweight Implementation of Per-packet Service Protection
by Jamal Hadi Salim 04 Oct '23

04 Oct '23

There are applications which MUST have reliable communication otherwise things fall apart! Think industrial automation, robotics, aerospace, automotive... Failure in timeliness of message delivery could be catastrophic. Typically when such a network is built, the operator makes sure that any congestion-related issues and packet losses are avoided. However, Murphy's law applies: the wiring may break, a radio link can be temporarily shielded by an obstacle, or worse the network nodes can be damaged... To deal with such losses, industry standardization has been introduced to provide per-packet service protection - these include Frame Replication and Elimination for Reliability(FRER) for Layer 2 Ethernet networks and the Packet Replication, Elimination, and OrderingFunctions(PREOF) forLayer3 IP/MPLS networks. Ferenc Fejes et al have implemented FRER in two different approaches: 1) in eBPF/XDP as a lightweight software implementation; and (2) in userspace using PF_PACKET. Ferenc et al will go over the architecture details of the two approaches and then evaluate XDP FRER via an experimental analysis and further compare the two FRER implementations. cheers, jamal Reminder: Three days to go before early-bird registration ends!

1 0

0x17: Talk, Synce4l: open-source implementation of Synchronous Ethernet
by Jamal Hadi Salim 04 Oct '23

04 Oct '23

Almost _a quarter_ of the world-wide consumer Internet traffic is pushed through mobile networks! Wireless mobile network traffic requires tight time synchronization with wired components of the 5G/6G network. PTP with its sub-microsecond synchronization just doesnt cut it anymore and Michał Michalik and Arkadiusz Kubalewski say we need sub-nanosecond synchronization! The only way we can achieve this is to use Synchronous Ethernet (SyncE) which allows transport of clock signals through the Ethernet physical layer. This prevents the clock skew and clock jitter between adjacent network nodes. In this talk, Michał and Arkadiusz will motivate us with the need of frequency synchronization and its role when coupled with PTP in 5G/6G networks. They will discuss the open-source project synce4l [0] and demonstrate, focusing mainly on the features, requirements (both in software and hardware), and configuration in a real network. Finally, the talk will describe how synce4l corresponds with the currently discussed new DPLL (Digital Phase Locked Loop) subsystem [1] in the Linux kernel and user-space networking community. [0] https://github.com/intel/synce4l [1] https://lore.kernel.org/netdev/20230824213132.827338-1-vadim.fedorenko@linu… cheers, jamal Reminder: Three days to go before early-bird registration ends!

1 0

0x17: Talk, Unstoppable Session Layer
by Jamal Hadi Salim 04 Oct '23

04 Oct '23

Back in circa 0x15[0], Alex Aring asked if you have an ECONNRESET on your socket, and even recovered, should we consider the communication reliable? Back then Alex made the argument that when an application encounters connection errors, the act of merely reconnecting reduces the likelihood of the application being considered reliable. He suggested a session layer to help in improving the reliability. To build up on that experience, Alex is back with what he calls "Unstoppable Session Layer" (USL) to allow you to easily switch your networking application to provide a session layer between application and transport layer. This talk will show how to easily put such a session layer on your networking application. It will be demonstrated by a basic TCP ping-pong application while tcpkill [1] tries to run crazy and interfere with the TCP application by sending TCP resets. The eventual goal is to port this implementation into the kernel so that that applications can benefit from it. [0] https://netdevconf.info/0x15/session.html?ECONNRESET-on-a-reliable-socket-a… [1] https://en.wikipedia.org/wiki/Tcpkill cheers, jamal Reminder: Three days to go before early-bird registration ends!

1 0

0x17: talk, Using eBPF to inject IPv6 Extension Headers
by Jamal Hadi Salim 04 Oct '23

04 Oct '23

IPv6 Extension Headers (EH) allows the IPv6 to be more flexible and extensible. Justin Iurman et al say, unfortunately, the EH does not survive easily on the global internet and it is often tricky to analyze these losses without real deployment; and for this reason they built an eBPF instrumented tool based that operators can use to test the survivability of specific IPv6 EHs. The tool can inject any of the extension headers (Hop-by-Hop Options Header, Destination Options Header, Routing Header, the Fragment Header, etc). The talk will go into the details of both the kernel and user space architecture and code of the tool. Justin will further demonstrate its potential usage through a use case deployed by a network service provider. cheers, jamal Reminder: Three days to go before early-bird registration ends!

1 0

0x17: talk, Quantum-Proofing Data: The Power Of Post-Quantum Cryptography
by Jamal Hadi Salim 04 Oct '23

04 Oct '23

Quantum computing presents a potential threat of being able to compromise existing cryptographic techniques. Shor's algorithm has the potential to break the RSA and Elliptic Curve Diffie-Hellman algorithms, both of which are pivotal for key exchange in Transport Layer Security (TLS). Our speaker, Milena Olech, says the solution is to adopt post-quantum cryptography (PQC) algorithms. In this talk Milena outlines the mathematical foundation of the CRYSTALS-Kyber algorithm and illustrates its application in the TLS handshake. In order to assess the efficacy of PQC, Milena will detail a comprehensive comparison between traditional cryptographic methods and post-quantum alternatives and provide a condensed summary of the comparison. cheers, jamal Reminder: Three days to go before early-bird registration ends!

1 0

Register now please!
by Jamal Hadi Salim 03 Oct '23

03 Oct '23

Just a reminder 5 more days to go for early bird registration. Also a reminder this is a hybrid conference; nothing can beat physical get together but you can participate virtually as well. Early registration helps us plan better. See: https://netdevconf.info/0x17/pages/registration.html cheers, jamal

1 0

0x17: talk, Integrating eBPF Into The P4TC Datapath
by Jamal Hadi Salim 03 Oct '23

03 Oct '23

P4TC was first introduced in 0x16[1]. After the first RFC was posted on the list, some of the community feedback made was to convert the s/w datapath from a scriptable (ala tc u32) to instead make use of eBPF with a strong argument being made that the s/w data path will be more performant. In this talk Jamal Hadi Salim et al will discuss the new architecture which factors in eBPF as the software datapath. The talk will go into the P4TC architecture factoring in eBPF and the pros and cons of this change from a scriptable approach. Jamal (twit that horn here) will further discuss the uapi and the kernel code architecture. There have been several RFC postings in the last 9 months[2] and it is expected the upstreaming efforts will start in the near future. cheers, jamal [1]https://netdevconf.info/0x16/session.html?Your-Network-Datapath-Will-Be-P4-Scripted [2] latest at this posting can be found here: https://lore.kernel.org/netdev/20230930143542.101000-1-jhs@mojatatu.com/

1 0

0x17: talk, Firewall and Service Tickets (FAST)
by Jamal Hadi Salim 02 Oct '23

02 Oct '23

There is no good standardized way for applications to signal the network for services, like QoS, that the app wants applied to their packets. Typically it is some sys admin that decides how these services and resources are partitioned and configures their control. Enter FAST - which is being developed at the IETF. Tom Herbert says FAST is secure, expressive, stateless, not spoofable, deployable, efficient, dynamic, requires no DPI, doesn’t disclose internals of the network, and works with any transport protocol. FAST tickets are attached to packets to describe the service requests and grants of admission that the application has requested from the network. Tickets are sent in IPv6 Hop-by-Hop options. Tom will provide rationale why Hop-by-Hop options are really the best choice compared to proposed alternatives despite concerns about them being dropped in the Internet. Eventually Tom will present the kernel patches and tie them to how applications will use the provided infra to request for tickets and how they are granted, how they're set and transmitted in packets, how a network node can process the FAST ticket in the packet and apply network mechanisms to satisfy the service requests. cheers, jamal PS: 5 more days left for early-bird registration. visit netdevconf.info/0x17/registration

1 0

[RESEND]0x17: Talk, Congestion control architecture for host congestion
by Jamal Hadi Salim 02 Oct '23

02 Oct '23

Apologies, I left out the subject line in the last email, so resending this... Link speeds are going up at an incredible rate. Cloud vendors are already asking for 4-800Gbps. Unfortunately the rest of the host hardware infra is not keeping up. Even at the "low low speed;->" of 100Gbps the host CPU speeds, cache sizes, memory access latency, memory bandwidth per core, NIC buffer sizes, are just overwhelmed. As an example 6 CPU cores can saturate all available memory bandwidth - imagine being on a machine that has 56C(112 hyperthreads).... In this talk, Saksham Agarwal, Arvind Krishnamurthy and Rachit Agarwal say that this has brought in new trends not seen before: congestion within the processor, memory and peripheral interconnects of the host which of course causes back pressure all the way to the NIC - meaning back to the network and, alas, causes congestion within the network. Testing with Linux DCTCP shows that host congestion by our speakers found as much as 1% packet drops at the host, 35-55% throughput degradation, and 120-5000x tail latency inflation! Agarwal et al want us to look closely at more than just network heuristics to resolve this emerging challenge. As an example: applications generating CPU-to-memory traffic and clogging the memory bus are currently not considered as part of the congestion control calculus at all - even though they are most certainly part of the culprits. In this talk our esteemed speakers introduce what they call "hostcc" as a response to these challenges. Hostcc makes about ~800 line kernel code changes. A little bit more details: https://netdevconf.info/0x17/6 For the real meat-and-potatoes come to the talk of course! cheers, jamal PS: 5 more days left for early-bird registration. visit netdevconf.info/0x17/registration

1 0

0x17: Talk,
by Jamal Hadi Salim 02 Oct '23

02 Oct '23

Link speeds are going up at an incredible rate. Cloud vendors are already asking for 4-800Gbps. Unfortunately the rest of the host hardware infra is not keeping up. Even at the "low low speed;->" of 100Gbps the host CPU speeds, cache sizes, memory access latency, memory bandwidth per core, NIC buffer sizes, are just overwhelmed. As an example 6 CPU cores can saturate all available memory bandwidth - imagine being on a machine that has 56C(112 hyperthreads).... In this talk, Saksham Agarwal, Arvind Krishnamurthy and Rachit Agarwal say that this has brought in new trends not seen before: congestion within the processor, memory and peripheral interconnects of the host which of course causes back pressure all the way to the NIC - meaning back to the network and, alas, causes congestion within the network. Testing with Linux DCTCP shows that host congestion by our speakers found as much as 1% packet drops at the host, 35-55% throughput degradation, and 120-5000x tail latency inflation! Agarwal et al want us to look closely at more than just network heuristics to resolve this emerging challenge. As an example: applications generating CPU-to-memory traffic and clogging the memory bus are currently not considered as part of the congestion control calculus at all - even though they are most certainly part of the culprits. In this talk our esteemed speakers introduce what they call "hostcc" as a response to these challenges. Hostcc makes about ~800 line kernel code changes. A little bit more details: https://netdevconf.info/0x17/6 For the real meat-and-potatoes come to the talk of course! cheers, jamal PS: 5 more days left for early-bird registration. visit netdevconf.info/0x17/registration

1 0

0x17: Talk, NIC offloads at Hyperscale: experience, new offloads and validation
by Jamal Hadi Salim 01 Oct '23

01 Oct '23

Willem de Bruijn says that lack of clear operating conditions, vague feature definitions and dearth of publicly available tests makes it hard to deploy at scale NIC offloads from multiple vendors. Issues seen in deployments range from inconsistency between devices, scalability limits, incompatibility with real production workloads, and simply hardware bugs; even a commodity offload as a checksum is not as well defined if you look across multiple vendor implementations. Willem will discuss why providing a way to define these offloads precisely, in a way that scales and avoids common bugs, while combining expertise from multiple vendors and users is the way to go. More importantly he believes that sharing all those details in an open and unencumbered format would help alleviate if not solve the current wild-west approach. The effort is currently housed under the auspices of the "OCP NIC Core Offloads specifications". Willem will go over some critical offloads, from established to novel ongoing work, such as pacing and inline crypto and show how they can fail in practice. And last but not least provide an overview of some of the testcases which have already been contributed to the kernel tree including tests for RSS, HW-GRO, TSO and others. cheers, jamal PS: 6 more days left for early-bird registration. visit netdevconf.info/0x17/registration

1 0

0x17: talk, Improving multi-phy and multi-port interfaces
by Jamal Hadi Salim 01 Oct '23

01 Oct '23

How do we deal with a single MAC connected to multiple PHYs - whether chained or in parallel - with multiple front-facing ports? What about when multiple front facing ports utilizing a single PHY or, fun, multiple PHYs? Maxime Chevallier we have some support for these scenarios today where we hide the PHY behind a netdev. The consequences are there is no way to config an individual PHY when multiple PHYs are present on a link - think SFP transceivers for example. Selecting which front-facing port to use is also limited. In this talk, Maxime will describe ongoing work to support these complex topologies, the challenges faced and expected improvements. cheers, jamal PS: 6 more days left for early-bird registration. visit netdevconf.info/0x17/registration

1 0

0x17: Talk, Is scaling eBPF easy yet? A small step to one server, but giant leap to distributed network.
by Jamal Hadi Salim 30 Sep '23

30 Sep '23

After deploying eBPF for over a year in a distributed service provider production environment Ivan Koveshnikov has come to the conclusion that there are a lot more challenges than simply slicing and dicing packets....What happens on "day 2" after the celebrations ends? In this talk Ivan will discuss important topics: development process, application lifetime management, configuration management and delivery, monitoring and observability, and scaling in the service providers' network. Using eBPF in such an environment adds extra spicing to the process challenges. Come and learn from Ivan's experiences! cheers, jamal Reminder: 6 more days left for early-bird registration. visit netdevconf.info/0x17/registration

1 0

0x17: Power-Aware Network Performance BoF
by Jamal Hadi Salim 30 Sep '23

30 Sep '23

Traditional host Network performance measurements are based on analyzing the correlation between the two classical metrics of throughput and latency against cpu utilization. This approach leaves out an important metric: Power utilization. The huge influx of AI infrastructure - which ends up consuming network resources - has brought much needed attention to power consumption as a variable in network infrastructure. If you consider that the operational cost of power is going up (significantly in some parts of the world with the ongoing crisis) we cannot ignore networking infrastructure contributions to not just the cost factor but also the environmental harm brought forth with high power use. Nabil Bitar, Jamal Hadi Salim and Pedro Tammela feel that Linux, which dominates data center deployments, deserves special attention. There is not much literature or shared experience in this space so they hope to inspire a discussion by sharing their experiences with the community: How would one go about measuring power utilization for network workloads? How do we correlate metrics such as perf, throughput, etc to the power utilized? How would one go about saving power while still achieving an application's stated end goals? If you would like to share your experiences or solutions on this important topic please reach out to me. The agenda is still open. cheers, jamal

2 2

0x17: BoF, Compiler-networking workshop
by Jamal Hadi Salim 28 Sep '23

28 Sep '23

If you thought, as a Linux networking person, that you only needed to sift through the truth^Whype on Rust, CXL, P4, ML/AI then you are mistaken. You also need to learn about compilers!.. Ok, just kidding. But seriously - compilers are the future! In the near future a lot of the networking infrastructure code will be compiler generated. In this exciting workshop, Felipe Alemeida, Tom Herbert, Bala Sambasivam, Neil Turton and Stephan Diestelhorst we will discuss how compilers are developing to accelerate network processing with new techniques, and how essential they are for the higher throughput and flexibility that datacenters need. They will discuss different abstractions for network problems that compilers can leverage to optimize networking and how different abstractions and languages transform to actual programs and instructions for optimal execution on next-generation network devices. Topics include: eBPF compilers, P4 compilers, MLIR, new intermediate representations, data flow analysis, multi-platform support, static analysis, and other advanced topics related to compilers and networking. If you would like to discuss something on this related topic then please reach out to Tom and/or Felipe to be considered for the agenda. cheers, jamal

1 0