Simple shell script to monitor your node uptime
This script, once launched in the background on your mina node server will parse the Mina logs at regular intervals to get uptime ticks.
Each loop will add a line to a simple csv file with timestamp for the tick and 24 hours splipping node uptime_value (from 0 (0% uptime) to 96 (100% uptime)).
The script handles the ftp upload of this file to a distant web server that can handle the graphical display of the node uptime results.
./uptime.sh [StarDate] [Interval] [Reinit] [Debug] [ftp_server] [ftp_login] [ftp_pass]
| Param | Description |
|---|---|
StartDate |
Date to get transaction from (YYYY-MM-DDTHH:MM:SS) (default 2022-05-01T00:00:00) |
Interval |
Repeat each Interval |
Reinit |
(0/1) Reinit Statistics from StartDate |
Debug |
(0/1) Enable Debug Mode |
ftp_server |
(ftp://server.com) FTP server adress to upload uptime csv file |
ftp login |
FTP login |
ftp passowrd |
FTP password |
./uptime.sh "2023-01-01T00:00:00" 21600 1 0 "" "" ""
The generate filename is based on uname -n
Example
mycomputer.csv
The content of the file will consist of a csv file with date,value for each line.
Example
Time,Score
2023-05-01 00:00:00,91
2023-05-02 00:00:00,96
2023-05-03 00:00:00,96
2023-05-04 00:00:00,92
In the exemple file, we can see the script parse the logs every 24h to get the 24h splipping ticks every 24h. On May, 1st, 91 uptime ticks have been successfully sent to the uptime server. It means that 5 ticks have been missed. On May, 2d, 96 uptime ticks have been successfully sent to the uptime server. Mina server uptime for this date will be 100%
The way the script work is really simple.
It uses filtered journalctl parsing to get the numbers of ticks for the last 24h every Interval seconds and then append the result to the csv file.
d=$(date '+%F %H:%M:%S')
uptime_csv_content="$d,$(journalctl --user -u mina -S "24 hours ago" | grep "Sent block with state" | wc -l)"
uptime_csv="./${filename}.csv"
...
echo $uptime_csv_content >> "$uptime_csv"
Specifying ftp_server, ftp login, ftp passowrd in the command line will send the updated csv file each Interval to an ftp server (in a way the file can be used by a simple web page to display the graph result).
Example : Send updated uptime file to ftp://webserver.com every 2 hours (7200 seconds)
./uptime.sh "2023-01-01T00:00:00" 7200 0 0 "ftp://webserver.com" "user" "pass"
You can just rely on the CSV file to check your uptime.
You can use the file locally or on a remote machine to handle graphical representation of your uptime.
Source for a simple web page project displaying graph based on CSV file is avalable here :
https://github.com/naamahdaemon/naamahdaemon.github.io
And an example of rendering here : https://naamahdaemon.github.io/