Scoped, nested, aggregated python timing library
StopWatch operates on a notion of "spans" which represent scopes of code for which we want to measure timing. Spans can be nested and placed inside loops for aggregation.
For example:
with sw.timer('root'):
for i in range(50):
with sw.timer('inner_task'):
do_inner_task(i)
StopWatch requires a root scope which upon completion signifies the end of the round of measurements. On a server, you might use a single request as your root scope.
StopWatch produces two kinds of reports.
- Aggregated (see _reported_values).
- Non-aggregated or "tracing" (see _reported_traces)
Aggregated reports have one value for each timer name. The above example would have 2 reports.
Tracing reports keep one trace every time we enter a scope. The above example would have 51 reports.
By default, StopWatch dumps aggregated report to stdout after the root scope completes and resets internal state.
Installation
pip install dbx-stopwatch
Usage
Basic Usage
import stopwatch
sw = stopwatch.StopWatch()
with sw.timer('root'):
for i in range(50):
with sw.timer('inner_task'):
time.sleep(0.1)
with sw.timer('outer_task'):
time.sleep(1.0)
print stopwatch.format_report(sw.get_last_aggregated_report())
yields
************************
*** StopWatch Report ***
************************
root 6206.833 (100%)
inner_task 50 5200.742 (84%)
outer_task 1 1002.351 (16%)
Tags:
Multiple reports
sw = stopwatch.StopWatch()
while request_in_queue():
r = get_request()
with sw.timer('request'):
process(request)
Tag your reports
while request_in_queue():
r = get_request()
with sw.timer('request'):
sw.addtag(r.get_endpoint_name())
sw.addslowtag('500ms', 0.500) # Tag only if request takes >= 500ms
process(request)
Report to your own backend
sw = stopwatch.StopWatch(
export_aggregated_timers_func=my_export_aggregated_timers,
export_tracing_func=my_export_tracing_func,
)
Contributing
Contributions are welcome. Tests can be run with tox. Lint with flake8 You'll have to agree to Dropbox's CLA.
Issues
If you encounter any problems, please file an issue along with a detailed description.