SZARP - System Zbierania Archiwizacji i Prezentacji danych
Copyright © 1991-2010 Witold Kowalewski Praterm Automatyzacja Ciepłownictwa
Copyright © 1995-2009 Praterm S.A.
Copyright © 1991-2016 various authors (see AUTHORS file)
Table of contents
- What is SZARP?
- What does SZARP mean? How to pronounce it?
- What are the features of SZARP?
- What is the origin of SZARP?
- On what platform is SZARP running?
- How to install SZARP?
- Where can I find documentation?
- How can I get SZARP sources?
- How to setup my first SZARP server?
- How can I get help?
1. What is SZARP?
SZARP is a full-featured SCADA system designed for monitoring slowly changing industrial processes, for example operations of urban heating systems. It is completely free software, published under terms of the GNU General Public License 2.0 (or any later version).
2. What does SZARP mean? How to pronounce it?
SZARP is an acronym for a Polish name: System Zbierania ARchwizacji i Prezentacji danych, which means System for Data Acquisition, Archiving and Presentation. Digraph "sz" in Polish is pronounced similarly to English "sh", so "SZARP" should be pronounced as a word "sharp".
3. What are the features of SZARP?
SZARP architecture is organized into three layers.
The first layer are device drivers. There are several drivers available, including:
- Modbus RTU/TCP (master and slave)
- Praterm ZET/Sterkom PLC
- Pollustat, Infocal and Kamstrup heating meters
- DDE (Dynamic Data Exchange) protocol
- communication with Python scripts through IPC
- importing data from text files and text output of programs
These device drivers are called line daemons (for historical reasons).
The second layer consist of data collecting daemons that read data from drivers, calculate average probes and formulas (called defined parameters) and finally save data to SZARP database. Probes are generally available with 10-seconds resolution, but database contains 10-minutes average values. There is also a web server available for serving parameters' values through HTTP.
In the third layer there are client programs, including:
- Raporter - a program for viewing current (10-seconds averages) values of parameters.
- Controller - a program for signalizing parameters' values irregularities.
- Ekstraktor - a program for exporting data from databases to spreadsheet files.
- Draw - a powerful tool for viewing and analyzing historical data.
More information about SZARP architecture is available in documentation.
4. What is the origin of SZARP?
SZARP was developed since 1991 (sic!) by Praterm company and was used in all heating systems (over 20) owned by Praterm and that much other systems all over Poland. In 2007 Praterm decided to publish almost all of the code under GNU GPL. Nowadays SZARP is developed as an Open Source project on GitHub.
5. On what platform is SZARP running?
SZARP is developed and deployed for Debian GNU/Linux, but should compile and run on any modern Linux distribution on i386, amd64 and armel/armhf architectures. Client applications should also run on Windows XP/Vista/7 machines with NTFS filesystem.
6. How to install SZARP?
File INSTALL contains installation instructions.
7. Where can I find documentation?
Unluckily most of the user documentation is available only in Polish. Sources of documentation are in resources/documentation/new/ directory and available online. Most of the new code is pretty-well documented (in English) using Doxygen.
8. How can I get SZARP sources?
9. How to setup my first SZARP server?
First install szarp-server and szarp-wx packages (for details see INSTALL):
# apt-get install szarp-server szarp-wx
In a configuration stage, you have to set 4-letter name of a configuration
(called configuration prefix), for example use "test". Answer "yes" to a
question about creating
Now look at the
/etc/szarp directory. There are two files:
parstart.cfg- this files defines what SZARP services should be run on system start. By default computer is considered as SZARP server only when the hostname equals to a configuration prefix - if that so, appropriate services will be run. This is probably not your case, so modify the file that it contains:
PARCOOK=1 MEANER3=1 SENDER=1 PARAMD=1 PSETD=0 XSLTD=0 SSS=0
szarp.cfg- this is a SZARP system configuration file. There are a lot of available options, but for now you only need to set configuration prefix. Modify the beginning of the file so it looks like this:
$if $prefix$="" $prefix$:="test" # $prefix$:=exec("hostname -s") $end
$prefix$ variable should be set to "test")
If all above is done, you have to prepare SZARP parameters configuration. We
will make configuration with one line daemon that polls system for average
load. Go to the directory
/opt/szarp/test and create subdirectory called
config. Place there a file named
params.xml with the following content:
<?xml version="1.0" encoding="utf-8"?> <params xmlns="http://www.praterm.com.pl/SZARP/ipk" xmlns:exec="http://www.praterm.com.pl/SZARP/ipk-extra" version="1.0" read_freq="10" send_freq="10" title="Test configuration"> <device daemon="/opt/szarp/bin/execdmn" path="/opt/szarp/test/get_la.sh" exec:frequency="10"> <unit id="1" type="1" subtype="1" bufsize="1"> <param name="Test:System:Average System Load" draw_name="Average Load" short_name="lavg" unit="-" prec="2" base_ind="auto"> <raport title="System report" order="1"/> <draw title="System" min="0" max="100" order="1"/> </param> </unit> </device> </params>
Then, in directory
/opt/szarp/test/config run command:
(this parses XML configuration, if every thing is ok, you should see "OK")
Prepare an executable script
/opt/szarp/test/get_la.sh with the following
#!/bin/bash cat /proc/loadavg | cut -d ' ' -f 1 | tr -d '.'
Save it and restart SZARP services:
# /etc/init.d/parstart restart
And that's all, you should be able to point your browser to http://localhost:8081/ and see your configured parameter. After full 10 minutes you can also run Draw3 program:
and analyze your system average load.
10. How can I get help?
You can try to write us at email@example.com.