Parsar

Parsar is a Python SAR parser designed to quickly and easily parse SAR data from plain-text SAR files.

It can be run from either the command line:

$ time parsar cpu sample.txt > /tmp/parsar.out
real    0m0.277s
user    0m0.268s
sys     0m0.004s

$ wc -l /tmp/parsar.out
7794 /tmp/parsar.out

Or in Python code:

>>> import parsar
>>> import time
>>> p = parsar.Parsar('sample.txt')
>>> def getcpustats():
...     start = time.time()
...     data = p.cpu()
...     end = time.time()
...     print 'Getting %d cpu stats took %.2f seconds.' % (len(data), (end - start))
...
>>> getcpustats()
Getting 7794 cpu stats took 0.23 seconds.

And produces CSV output that is easy for graphing tools to consume:

$ head -n 10 /tmp/parsar.out
#Time,%idle
2017-05-08T21:11:09,79.92
2017-05-08T21:11:12,89.86
2017-05-08T21:11:15,77.42
2017-05-08T21:11:18,63.74
2017-05-08T21:11:21,69.48
2017-05-08T21:11:24,44.79
2017-05-08T21:11:27,55.61
2017-05-08T21:11:30,60.89
2017-05-08T21:11:33,74.08

Supports returning whatever data column(s) you're interested in:

$ parsar cpu --cpustats %usr %iowait sample.txt > /tmp/parsar.out
$ head -n 10 /tmp/parsar.out
#Time,%usr,%iowait
2017-05-08T21:11:09,3.95,13.25
2017-05-08T21:11:12,3.90,5.41
2017-05-08T21:11:15,6.93,13.02
2017-05-08T21:11:18,4.87,26.74
2017-05-08T21:11:21,9.15,19.98
2017-05-08T21:11:24,17.04,34.81
2017-05-08T21:11:27,23.35,16.41
2017-05-08T21:11:30,3.48,32.10
2017-05-08T21:11:33,6.13,18.62