Future Tech

Tencent Cloud's home-grown traffic-tamer halves latency

Tan KW
Publish date: Mon, 05 Aug 2024, 04:51 PM
Tan KW
0 462,597
Future Tech

Sigcomm 2024 Chinese web giant Tencent has revealed MegaTE - a traffic engineering system it uses on its own cloud and which it claims outperforms rivals by tailoring network performance to the needs of individual flows.

A paper detailing MegaTE was presented today at the Association for Computing Machinery's SIGCOMM conference in Sydney, Australia, by Tencent senior researcher Congcong Miao. He explained that virtual machines in cloud environments typically interact with a bandwidth broker that finds an optimal path and tries to assign the most appropriate resources to get traffic moving.

That's not efficient in a hyperscale cloud, he asserted, because applications need networks to cater to their particular needs - either due to the nature of the workload or the service levels promised by a cloud operator.

Today's traffic engineering (TE) tools - Miao named scFLOW and TEAL (Traffic Engineering Accelerated by Learning) - aren't designed to cater to the needs of each flow.

MegaTE, by contrast, can "satisfy the needs of each fine-grained traffic flow at the virtual instance level."

Miao explained the tool does not assume it is possible to arbitrarily decide what traffic flow between two endpoints will need. Instead, endpoints run an agent that takes advantage of the Berkeley Packet Filter (eBPF) - tech that allows code to run in sandboxes within the Linux kernel. That agent collects traffic to create flow data that is eventually shared with a controller that calculates an optimal network path. That calculation is sent back to the agent, which stores it in an "eBPF map" so the application's needs can be advertised and acted upon.

MegaTE is also aware of network topologies.

Packets leaving a cloudy endpoint include routing information, meaning that their passage across a network has already been planned and should be smoother.

The result, Miao told the conference, is that high-priority applications can meet service level agreements. Tencent uses it in production and apparently reduced packet latency across the WAN by 51 percent. Miao claimed the tool achieved that while handling over 20,000 flows at a time - many more than is possible with either scFLOW or TEAL.

Miao noted that the metrics he mentioned reflect performance of Tencent's own workloads, rather than outcomes for Tencent Cloud tenants.

But the savings he mentioned will likely be very significant for Tencent. Its WeChat and Weixin services have over 1.5 billion users and its video and music streaming services boast 100-million-plus subscribers apiece. The concern also runs China's second-biggest app store, and is a colossal games publisher and operator. Tencent's Cloud spans seven regions and 31 availability zones across China and Hong Kong, plus another eleven regions and 22 availability zones in other countries.

The giant likely operates millions of servers and a substantial WAN to connect them all - making MegaTE's potential impact enormous. ®

 

https://www.theregister.com//2024/08/05/tencent_clouds_homegrown_traffictamer_halves/

Discussions
Be the first to like this. Showing 0 of 0 comments

Post a Comment