Skip to content

NLnetLabs/simdzone

Repository files navigation

Build Status Coverity Status Mastodon Follow

simdzone: Parsing zone files really fast

Fast and standards compliant DNS presentation format parser.

DNS resource records (RRs) can be expressed in text form using the presentation format. The format is most frequently used to define a zone in master files, more commonly known as zone files, and is best considered a tabular serialization format with provisions for convenient editing.

The format is originally defined in RFC1035 section 5 and RFC1034 section 3.6.1, but as the DNS is intentionally extensible, the format has been extended over time too.

This project provides a lightning fast presentation format deserializer (and serializer eventually) for other projects to leverage. Learn more about simdzone by reading the documentation.

Motivation

Zone files can become quite large (.com ~24G, .se ~1.3G) and the parser in NSD left something to be desired. simdjson demonstrates that applying SIMD instructions for parsing structured text can significantly boost performance. simdzone, whose name is a play on simdjson, aims to achieve a similar performance boost for parsing zone data.

Currently SSE4.2 and AVX2 are supported, a fallback is used otherwise.

simdzone copies some code from the simdjson project, with permission to use and distribute it under the terms of The 3-Clause BSD License.

Results

Running zone-bench on my system (Intel Core i7-1065G7) against an older .com zone file of 12482791271 bytes under Linux (Fedora 39).

clang version 17.0.6, release mode:

$ time ./zone-bench parse ../../zones/com.zone
Selected target haswell
Parsed 341535548 records

real    0m13.533s
user    0m12.355s
sys     0m1.160s

There are bound to be bugs and quite possibly smarter ways of implementing some operations, but the results are promising.

Compiling

Make sure the following tools are installed:

  • C toolchain (the set of tools to compile C code)
  • cmocka (if configured with -DBUILD_TESTING=on)
  • Doxygen (if configured with -DBUILD_DOCUMENTATION=on)
  • Sphinx (if configured with -DBUILD_DOCUMENTATION=on)

To compile in release mode:

$ cd zone-parser
$ mkdir build
$ cd build
$ cmake -DCMAKE_BUILD_TYPE=Release ..
$ cmake --build .

To compile in debug mode with testing:

$ cd zone-parser
$ mkdir build
$ cd build
$ cmake -DCMAKE_BUILD_TYPE=Debug -DBUILD_TESTING=on ..
$ cmake --build .

Contributing

Contributions in any way, shape or form are very welcome! Please see CONTRIBUTING.md to find out how you can help.