Internal Logger Documentation

Logger is a component that functions to record events or activities in a system/application. It is usually used for system monitoring and debugging.

Why does Storm have a logger?

The logger in Storm is custom, and we optimized the code to achieve speed and stability against massive I/O, and memory safety.

We made it using the Rust programming language because Rust is a low-level language that guarantees speed equivalent to the C programming language but with very good memory safety.

Logger Features

Storm Logger has 2 different logic mechanisms in handling a system log.

1. smf.printf: It is a standard log function that will output to the terminal quickly.
2. smf.printd: This is the log for debugging, we made it not throw the log to the terminal but save the log to a local database SQLite in lib/sqlite/logging/.

Further explanation

smf.printf: This is designed to capture Logs very efficiently compared to using Python's built-in print which is a bit of a bottleneck and has a bit of a delay although the comparison is slightly different.

smf.printd: This log base is the same but only it captures logs with more detailed and complete data for efficient debugging and monitoring of Storm system behavior.

Debug Level

We use the debug level as usual, namely:

DEBUG, INFO, WARN, ERROR, CRITICAL: For filters to make it easier to find fatal logs.

Logger Usage

1. Use direct import into Python.

import smf

2. Then use the function:

smf.printf("example", err, file=sys.stderr, flush=True)
smf.printf("example", a, b, c)

smf.printd("example", err, level="ERROR")
smf.printd("example", a, b, c, level="DEBUG")

Export Log

Because for log smf.printd By default, logs are always saved in SQLite, so we need to export the log with the following command:

smf> export log <value>

content <value> with the appropriate log level, it will export the log according to the log level to a txt file and save it in $HOME.