6.1 KiB
libpsyc Performance Benchmarks
In this document we present the results of performance benchmarks of libpsyc compared with libjson-glib and libxml2.
Procedure
We'll use typical messages from the XMPP ("stanzas" in Jabber lingo) and compare them with equivalent JSON encodings, verbose and compact PSYC formats.
In some cases we will additionally compare PSYC packets to a more efficient XML encoding based on PSYC methods, to have a more accurate comparison of the actual PSYC and XML syntaxes, rather than the protocol structures of PSYC and XMPP.
The Benchmarks
A presence packet
Since presence packets are by far the dominant messaging content in the XMPP network, we'll start with one of them. Here's an example from paragraph 4.4.2 of RFC 6121.
And here's the same information in a JSON rendition:
Here's the equivalent PSYC packet in verbose form (since it is a multicast, the single recipients do not need to be mentioned):
And the same in compact form:
:c psyc://example.com/~juliet
=da 4
np
|An average chat message
XMPP:
JSON:
PSYC:
Why PSYC doesn't have an id? Because packet counting from contexts and circuits is automatic: The packet already has a number just by being there.
Also, PSYC by default doesn't mention a "resource" in XMPP terms, instead it allows for more addressing schemes than just PSYC.
A new status updated activity
Example taken from http://onesocialweb.org/spec/1.0/osw-activities.html You could call this XML namespace hell:
http://activitystrea.ms/head/json-activity.html proposes a JSON encoding of this. We'll have to add a routing header to it.
http://about.psyc.eu/Activity suggests a PSYC mapping for activity streams. Should a "status post" be considered equivalent to a presence description announcement or just a message in the "microblogging" channel? We'll use the latter here:
A message with JSON-unfriendly characters
A message with XML-unfriendly characters
A message with PSYC-unfriendly strings
A packet containing a JPEG photograph
… TBD …
A random data structure
In this test we'll not consider XMPP at all and simply compare the efficiency of the three syntaxes at serializing a typical user data base storage information. We'll again start with XML:
In JSON this would look like this:
Here's a way to model this in PSYC:
Results
Parsing time of 1 000 000 packets in milliseconds:
| input: | PSYC | JSON | XML | ||||
|---|---|---|---|---|---|---|---|
| parser: | strlen | libpsyc | json-c | json-glib | libxml sax | libxml | rapidxml |
| presence | 30 | 246 | 2463 | 10197 | 4997 | 7557 | 1719 |
| chat msg | 41 | 320 | 5997 | 9777 | 1893 | ||
| activity | 42 | 366 | 4666 | 16846 | 13357 | 28858 | 4419 |
| user prof | 55 | 608 | 4715 | 17468 | 7350 | 12377 | 2477 |
| / | < | > | < | > | < | > |
These tests were performed on a 2.53 GHz Intel(R) Core(TM)2 Duo P9500 CPU.
Conclusions
… TBD …
Criticism
Are we comparing apples and oranges? Yes and no, depends on what you need. XML is a syntax best suited for complex structured data in well-defined formats - especially good for text mark-up. JSON is a syntax intended to hold arbitrarily structured data suitable for immediate inclusion in javascript source codes. The PSYC syntax is an evolved derivate of RFC 822, the syntax used by HTTP and E-Mail, and is therefore limited in the kind and depth of data structures that can be represented with it, but in exchange it is highly performant at doing just that.
So it is up to you to find out which of the three formats fulfils your requirements the best. We use PSYC for the majority of messaging where JSON and XMPP aren't efficient and opaque enough, but we employ XML and JSON as payloads within PSYC for data that doesn't fit the PSYC model. For some reason all three formats are being used for messaging, although only PSYC was actually designed for that purpose.
Caveats
In every case we'll compare performance of parsing and re-rendering these messages, but consider also that the applicative processing of an XML DOM tree is more complicated than just accessing certain elements in a JSON data structure or PSYC variable mapping.
For a speed check in real world conditions which also consider the complexity of processing incoming messages we should compare the performance of a chat client using the two protocols, for instance by using libpurple with XMPP and PSYC accounts. To this purpose we first need to integrate libpsyc into libpurple.
Futures
After a month of development libpsyc is already performing pretty well, but we presume various optimizations, like rewriting parts in assembler, are possible.
Appendix
Tools used
libpsyc:
test/testStrlen -sc 1000000 -f $file test/testPsycSpeed -sc 1000000 -f $file test/testJson -snc 1000000 -f $file test/testJsonGlib -snc 1000000 -f $file
xmlbench:
parse/libxml-sax 1000000 $file parse/libxml 1000000 $file parse/rapidxml 1000000 $file