User guide

The following user guide describes how to use the Hyrise Cockpit.
Die folgende Anleitung beschreibt wichtige Prozesse in der Bedienung des Hyrise-Cockpits.

Add a database

Press the field to display the database list:
Press the button to add a database:
Enter the data. The database id is used directly by the host. You can still manually change it to your preference. How exactly the host must be specified is determined by the network configuration. For example, if a Hyrise instance is running on the machine 123.123.123.123 on port 1234, you would need to add 123.123.123.123 in the host field and 1234 in the port field. If you added the host and port you can also select the number of workers. A worker is a process that sends the queries from a workload to the hyrise. More workers mean more pressure on the hyrise. But be aware that workers cost system resources and sometimes just a few workers are enough. After you selected the number of workers click on the 'Save' button:

Delete a database

Press the field to display the database list:
Click the delete symbol:
Click the 'Remove' button:

Send an SQL query

Press the field to display the database list:
Click the Sql-Button:
Enter the SQL query and click "send query" to send it to the database:
After the SQL query is processed you can see the results of it:

Activate and deactivate plugins

Press the field to display the plugin overview:
Activate or deactivate the desired plugin by pressing the switch (The plugin window can be moved freely):

Chanage settings of a plugin](#change-settings-of-a-plugin

Press the field to display the plugin overview:
By pressing the gearwheel when the plugin is activated, all possible settings are displayed:
You can change the plugin settings and click 'Save'. Moreover, you can get an explanation for all the settings by hovering over the 'question mark':

Get the plugin-Logs inside of the plugins overview

Press the field to display the plugin overview:
By pressing on 'plugin log messages' the display can be expanded:
The text field of the logs updates itself and can be scrolled:

Get the plugin-Logs inside of the diagrams

Log messages are drawn as gray bars in the line graphs at the time they occur:
If you click on these lines you get the log message displayed. (You can move these log messages freely and can close them by pressing 'x'):

Loading and unloading of data into the database

Click the field for the workload settings:
Select in the workload setting the data you want to load. The data consists of the binary tables for the selected workload/benchmark:

Start a workload

Click the field for the workload settings:
First you need to load the data (binary tables) for the workload you want to run. If you want to run a mix of workloads (for example TPC-H and TPC-DS) you need to load for every workload the data. Otherwise, no binary tables are loaded and a start is not possible:
Select the workload (or workloads) and define the number of requests per second that should be sent to every database. For that, you can enter a number or change the slider. Subsequently, click on 'Play':
If a workload is running the 'Play'-Button is green:

End a workload

Click the field for the workload settings:
If the 'Play'-Button is green (so a workload is running) you can press stop (Delete al queries in the waiting queue):

Change a running workload

Click the field for the workload settings:
The SQL-Queries which are sent to the database per second can be changed by using the slider or text input. :
When the workload is activated, the detailed settings can be called up by switching tabs:
In the detailed settings, a weight can be set for each request in the workload using the slider or text input: Be aware, if you started your workload with a high frequency it is possible that the queue with all the queries that the worker uses to send the request to the database is very large. The new distribution of the queries defined by the new workload setting will take action until all queries from the time before the change of the settings are processed by the worker. All queries a process in LIFO. If you want to let the new distribution take immediate effect you need to restart the workload.

Edit the displayed metrics

Depending on which monitoring page (Overview, Comparison, and Workload Analysis) is currently open, it can be edited by pressing the gear:
For the respective page, the metrics can be selected and deselected by pressing the 'eye':

Edit the displayed databases

Depending on which monitoring page (Overview, Comparison, and Workload Analysis) is currently open, it can be edited by pressing the gear:
For the respective page, the databases can be selected and deselected by pressing the 'eye':

Setting the continuous-time and update intervals

Depending on which monitoring page (Overview, Comparison, and Workload Analysis) is currently open, it can be edited by pressing the gear:
Select the 'Time Range' tab:
The observed period can be set to the desired value using the 'Observed Time Range' dropdown: *The update interval can be set through the 'Aggregation Interval' dropdow (the possible intervals differ depending on the observed period):
If a static query is currently being carried out, then 'SET CONTINUOUS TIME RANGE' must be pressed (otherwise, the update happens automatically):
The green point behind 'Continuous Range' shows whether a continuous query of the database data is currently taking place ( Continuous means that the data is updated at regular intervals):

Setting static time periods

Depending on which monitoring page (Overview, Comparison, and Workload Analysis) is currently open, it can be edited by pressing the gear:
Select the 'Time Range' tab:
Switch to the page for editing static periods below:
Specify start and end date and time (these must be in the past):
Specify the interval for aggregating the points:
click 'SET STATIC TIME RANGE':
The green point behind 'Static Range' shows that static diagrams are currently being displayed (static diagrams do not update automatically)