Installation

The AutoED package is available through Python’s package manager. To install it, run

pip install autoed

Use autoed -h to see its full options list. The main idea behind the package is to monitor directories with diffraction images, to register and process new files obtained from the eBIC microscope.

Usage

First, start the daemon process.

autoed start

Next, start watching a directory.

autoed watch /path/to/diffraction/data

This will create a subprocess that runs separately from the AutoED daemon. The watchdog process monitors all subdirectories within the given path (recursively).

You can list currently watched directories with

autoed list

which will give you a list of all the watchdog processes, their process identifier number (PID), and the path to the directory being watched.

Listing watched directories
PID    PATH
------------------------------------------
430056  /home/.../.../dir1
882356  /home/.../.../dir2
------------------------------------------

* Please use 'autoed kill PID' to kill a process.
  Alternatively you can use 'pkill -P PID'.

As the output says, you can terminate a watchdog script using its PID. For example,

autoed kill 430056

to stop watching the first directory. Alternatively, you can run

autoed stop

to kill all the watchdog processes and terminate the autoed daemon.

Important

AutoED uses inotify to monitor the filesystem. Unfortunately, inotify does not work on Network File Systems (NFSs). Because of this, AutoED uses the polling method in inotify. The disadvantage of the polling method is that it might require more computational resources. If you are working on a local filesystem, you can run the watch command with an option --inotify or -i to use inotify without polling. By default, AutoED will use the polling method.

To minimize the CPU usage, you can set the time interval (in seconds) for filesystem checks. Use --sleep_time or -t. For example,

autoed -t 60 watch /path/to/diffraction/data

would check the filesystem every minute.