News for hackers. Stuff that matters.
Re:SocketCluster – WebSockets that scale to 100K messages pe... by teacup50 on Wednesday May 07, 2014 @07:41PM (#7713217)
Is 100K mps on 8 cores considered high for node/websockets microbenchmarking of the socket path?

That doesn't seem like much from past experience writing high-throughput messaging code, and all this is doing is spitting out length-framed messages to a socket.

Re:SocketCluster – WebSockets that scale to 100K messages pe... by bhauer on Wednesday May 07, 2014 @09:04PM (#7713459)
We do not (yet) measure the performance of Websocket in our project (the TechEmpower framework benchmarks), but our "Plaintext" test is a rough analogue to a ping-pong test on Websocket. Our Plaintext test uses HTTP pipelining on a keep-alive connection. However, in our case, each request is sending a couple hundred bytes of HTTP request headers and receiving about the same in response HTTP headers prior to a "Hello world" payload.

We see approximately 600,000 of these HTTP "messages" per second on a i7-2600K 8 HT core workstation [1] from top performers such as Netty and Undertow and these top performers are network limited by our gigabit Ethernet.

We are using Undertow for a Websocket project presently and its performance there has been very good.

[1] http://www.techempower.com/benchmarks/#section=data-r9&hw=i7...

Re:SocketCluster – WebSockets that scale to 100K messages pe... by bilbo0s on Wednesday May 07, 2014 @07:59PM (#7713280)
It's not high at all.

I have never worked with node though... so maybe it's ... sort of ... "fast for node". Not fast in absolute terms.

Re:SocketCluster – WebSockets that scale to 100K messages pe... by Pacabel on Wednesday May 07, 2014 @08:00PM (#7713284)
That's what I was thinking, too. That's merely around 13,000 messages per second per core. Rates like that weren't all that impressive on low end server hardware a decade ago, so I'd hope that it's even less impressive today when using more modern hardware (or even VMs).
Re:SocketCluster – WebSockets that scale to 100K messages pe... by zoomerang on Wednesday May 07, 2014 @10:09PM (#7713624)
I'd be expecting more in the realm of millions of messages per second, at least from a Java solution.
Re:SocketCluster – WebSockets that scale to 100K messages pe... by limsup on Wednesday May 07, 2014 @07:58PM (#7713278)
Can you give some numbers of your past experience?
Re:SocketCluster – WebSockets that scale to 100K messages pe... by brianwawok on Wednesday May 07, 2014 @08:30PM (#7713358)
Not OP, but a solar flare can shoot 20 million packets per second..

http://www.solarflare.com/09-11-12-Solarflares-Industry-Lead...

You need to totally bypass the linux kernel to do this, and have high perf C or Java code...

Re:SocketCluster – WebSockets that scale to 100K messages pe... by chmod775 on Wednesday May 07, 2014 @08:51PM (#7713425)
That's raw network (TCP/UDP) packets, not WebSocket frames/messages. Plus that product is a solution to a hardware problem, not a software problem. It doesn't relate at all.
Re:SocketCluster – WebSockets that scale to 100K messages pe... by pktgen on Thursday May 08, 2014 @04:54AM (#7714452)
Just for anyone else who sees that and thinks "WTF": it's 20M PPS across two ports. (14,880,952 is the max for 10GbE.)