NovaStar Program Reference / Data Analysis / tstool
Overview
The third-party TSTool program, run with the tstool
script, processes time series and other data into useful data products.
The benefits of using TSTool include:
- it provides extensive time series and data table processing functionality, including ability to process data from various sources and formats
- data processing can be prototyped outside of the server and then automated as a scheduled process if needed
- data processing load can be offloaded from the server
The program can be run as a scheduled process on a frequency consistent with data collection cycle. TSTool can also be run interactively on Linux or Windows computers and communicates with NovaStar REST web services.
TSTool is developed and maintained by the Open Water Foundation and is installed by TriLynx Systems in support of NovaStar systems. See the following for more information:
- TSTool product page at Open Water Foundation. Document is currently available as navigable PDF once the software is installed (see TSTool Help menu).
Configuration
TSTool accesses the NovaStar system via a plugin developed by TriLynx Systems. A datastore is configured to access NovaStar REST web services to read data.
TODO smalers 2017-08-07 Steve needs to document more completely - need to determine best home for documentation.
Command Line Usage
The command line syntax on a NovaStar serer is as follows (if installed on Windows, the program can be run from the Start menu):
/opt/TSTool-latest/bin/tstool [parameters]
If run with no parameters, the interactive graphical user interface will be displayed.
Use the -commands commandFile
parameter to specify a command file and run in batch mode.
Command line parameters are:
Parameter | Description | Default |
---|---|---|
-commands commandFile |
Process a TSTool command file. | Command file is not used, interactive mode is used. |
Examples
Run Flood Impact Tool
TSTool is used to implement the Flood Impact Tool.
In this case TSTool reads instantaneous Discharge-River
data type time series at gage locations,
converts the data to regular-interval time series, and routes the flows to asset locations.
The normal convention is to create a command file on the server and run TSTool as a scheduled process
so that routed flow time series can be read by the NovaStar REST web services /floodImpactSummaries
service.
The command file for processing can be tested and then installed on the server to run as a scheduled process.
The following process can be scheduled in the Administrator Interface to run the Flood Impact Tool routing:
/opt/TSTool-latest/bin/tstool -commands /var/lib/novastar5/resources/FloodImpactTool/FloodImpactTool.TSTool
Administrator Interface
TODO smalers 2017-07-06 need a screenshot for a scheduled process
Log File Location
TSTool creates a default log file in the .tstool/log
folder under the user account that is running the program.
Other log files are typically created in the working folder for a process by using the TSTool StartLog
command.
Performance Considerations
The following should be considered to optimize TSTool performance:
- TSTool is a large Java application and can take several seconds to start up.
Rather than running many TSTool processes, TSTool's functionality should be used to process multiple products.
For example, use the TSTool
For
command to loop over a list of stations read with theWebGet
andReadTableFromDelimitedFile
or similar commands. - Test performance on a desktop computer before automating server-side processes.
Troubleshooting
Problems are typically due to data or command workflow issues. Running TSTool interactively will display markers on commands that have issues and messages can be viewed interactively by right-clicking on a command and selecting Show Command Status. Also refer to the log file using the Tools / Diagnostics - View Log File menu.