How is it possbile to see the used sizes of the storage that is used by a AWS Lamdba function?
For example, like this it is possibe to check if a file exists:
import os
os.path.isfile('/tmp/' + filename)
However, I need to know when when the 512 MB limit is reached.
Since the Lambda Function is just running on a container in a linux environment, we can use the OS to tell us generally how much space we have left on the temporary filesystem.
from subprocess import check_output
out = str(check_output(["df", '-k']))
print(out)
result = re.search(r"(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\%\s+\/tmp", out)
size, used, available, used_pct = (int(g) for g in result.groups())
print(size, used, available, used_pct) # 538424 440 526148 1
Here's what's happening here. First We use the subprocess.check_output function to run the command: df -k which displays a string about information on the disk space usage in KB. Here's a sample of that output on a lambda function:
Filesystem 1K-blocks Used Available Use% Mounted on\n/dev/root 6127168 4978732 1132052 82% /\n/dev/vdb 1965904 45296 1904224 3% /dev\n/dev/loop0 538424 440 526148 1% /tmp\n
You can see that at the end of the string is there the /tmp partition is located. So we take the output and search it with the regex pattern: (\d+)\s+(\d+)\s+(\d+)\s+(\d+)\%\s+\/tmp which separates out the relevant fields into 4 groups which we can set to their own variables that represent the size: total partition size (in KB), used: The amount of space currently being used (in KB), available: The amount of space still available on the partition (in KB), and used_pct: The percentage of the partition space that is being used.
This is a pretty simple way of getting at the space available, I would not use this if precise measurements are needed or for guessing if a new file can be written to the disk. But if you're just logging it on execution for debugging, it should be perfectly fine.