Today I spent quite some hours in spotting some packet drops at Gigabit line rate. Stress-testing the hardware at line-rate reveal a lot of noise and disagreements: it is difficult to determine _what_ really is the source of packet drops. Due to some netperf and perf analysis and some new tracepoints the eyes focus on PCI express bus. Quite a lot of analysis until that, first I interpreted the raw values as a CPU limitation....
Last time as I posted about TCP fairness aspects of TCP I don’t touched the topic in great detail. But TCP fairness and protocol fairness in general is a important requirement in the whole IP communication zoo. In this blog posting I will write some sentences about the fairness aspects of a network layer protocol and particular TCP. UDP and Datagram based Protocols To start with UDP - as a unreliable, unordered lightweight datagram based service - provides no fairness mechanisms at all....
Traveling back from Berlin to Munich. This time via City-Night-Train booked in a six-person wagon. Reading “der Freitag”, a Berlin weekly newspaper produced here in Berlin until midnight. One of Kitas professor suggest the paper. I took this extended weekend (from Thursday) to arrange our new home in Weissensee/Berlin. Paint the rooms white and the floor in a special mixed color. Next time I will establish the parquet (84m^2). After that all ground work is done and the furniture can be ebayed[TM]....
Between 15th and 25th of July I was at Majorca/Spain for one week of relaxing (including a little bit of climbing). A really nice spot!
Quit interesting Westmere-EP CPU: of cores | 6 | of threads | 12 | Bus speed | 2933 MHz QPI | System Bus|5.86 GT/s| Level 1 cache size | 6 x 32 KB icache, 6 x 32 KB dcache | Level 2 cache size | 6 x 256 KB | Level 3 cache size | 12 MB | Features | SSE4.2, Turbo Boost, AES New Instructions, Enhanced Intel SpeedStep Technology. … | Max Memory Bandwidth | 32 GB/s| Memory Types| DDR3-800/1066/1333| Lithography| 32 nm|...
“Libhashish”:http://libhashish.sourceforge.net/ is one of my long term project developed with git. Today I stumbled over a tool to visualize commits, called gource: Your browser does not support the video tag (or webm codec).
“Mobility Support in IPv6”:http://tools.ietf.org/html/rfc6275 - Request for Comments: 6275
Jesper Juhl today on LKML questioning about the new Kernel naming scheme: “… There are a few scripts that need fixing if SUBLEVEL goes the way of the Dodo, but I didn’t want to start fixing those without a clear indication of whether or not SUBLEVEL is going to/should die.”
.h2 Move VisualSimulatorImpl into the core module? move VisualSimulatorImpl into the core or network module. Does this remind anyone else of Microsoft moving graphics into the kernel in NT 4.0, so that it was much more unreliable than NT 3.5 and could be crashed by any 3D screensaver? -1
Tom posted these days a patch called “Allow no-cache copy from user on transmit”. The idea behind is the following: in net/ipv4/tcp.c:tcp_sendmsg() will copy date from user buffer to sk_buff via copy_from_user() or csum_and_copy_from_user() (which does what the name implies: copy the data and calculate a (partial) checksum). Tom realized that when data is copied data caches are touched which are often not used afterwards. The feature is enabled if the device signals that he is doing some kind of checksum offloading....