120 likes | 208 Vues
CZAR 2.0 is a robust system offering configuration management, data logging, and monitoring capabilities. It features advanced tools for history data retrieval, configuration control, and more. The system is highly scalable and customizable, with support for various databases and data compression methods. Currently in development, CZAR 2.0 promises enhanced performance and flexibility for data-intensive operations.
E N D
CZAR 2.0 (in development) Christopher A. Larrieu Chris Slominski
Components • Configuration Database • Data logging engine • Data store (Embedded Database) • History data server • Configuration management GUI • Engine control / monitor GUI • History data retrieval clients • Command-line control/config tools
System Overview Oracle 9i MySQL 4 XML Configuration Manager OCI Xerces CORBA DAQ Engine C++ History Server CORBA CORBA Manager GUI Data Store XARR CORBA Berkeley DB StripTool Command-line Tools Filesystem hapiget DARTH
New Stuff • Oracle or MySQL for configuration database • Berkeley DB for data store • Embedded database • Reasonable open-source license • Import legacy data • Data compression (6 TB / year .8 TB / year) • 84% lossless • 92%, truncating timestamps to milliseconds • Improved command-line tools • Expiration dates for logging requests
History Clients • XARR • Pure CORBA client • hapiget command-line tool • Script applications • StripTool • CORBA client in separate thread • DARTH (Java) • Extensive configuration • Flexible display (graph, table, correlation)
Status • Version 1.0 • operational since January 2002 • ~ 23,000 signals • ~ 3.6Gb / day raw; processed: 1Gb • HP-UX 11 Solaris 8 in June 2002 • Version 2.0 • Solaris 8 Linux 7.3 July 2003 • ~ 3.6Gb / day raw; processed: ~360Mb
More Status • Import config files from ChannelArchiver, old AR • New configuration and install scripts • > emacs config/prefs/myprefs.mk • > make build; make install • Init scripts for starting / stopping / inspecting entire system in standard unix fashion • > rc.czar start • HP-UX 11 aCC, Solaris 8 CC, Linux (g++ 3.2) • ACE/TAO 5.2.1 • Berkeley DB (SleepyCat)
Transportability • Entire system • Engine • Data Store • History server • CADataStore API • CORBA API
Scalability • Data store decoupled from data acquisition • multiple engines can spool data • Embedded database indexes, summarizes data • fast lookup • easily shuffle data offline/online • fast retrieval • History server • reduce data before transporting
In the Works • Data transformation API • Darth 1.0 • Public History API specification, IDL • Insertion API for externally acquired data