From: Ondrej Lichtner olichtne@redhat.com

Extending the PerfResult object api with the following methods/properties:

* start_timestamp * end_timestamp * time_slice(start, end)

These unify the api for PerfInterval, SequentialPerfResult and ParallelPerfResult.

This also means that "timestamp" itself in PerfInterval is renamed to "start_timestamp" to avoid duplication.

The "time_slice" method returns an object of the same type (PerfInterval, SequentialPerfResult, ParallelPerfResult or other class derived from PerfResult) that is "sliced" to only contain measurement data from the selected timestamp period restricted by the "start" and "end" parameters.

The implementation depends on the specific class: * PerfInterval serves as the smalles unit and creates a new PerfInterval object that has a shorter *duration* (intersection of current start-end and requested start-end), and has the *value*, adjusted by the ration of new_duration/old_duration. This ensures that the "average" calculated from this PerfInterval is consistent after the time_slice * PerfList derivatives (SequentialPerfResult and ParallelPerfResult), recursively call "time_slice" in all the individual items of the sequence.

In case an empty slice would be created (mismatch of start-end intervals), an EmptySlice exception is thrown.

This commit also adjusts the PerfRecipe "align_data" related code to use this new refactored code.

v2: * fixed Perf.Recipe time_aligned_results method - in case alignment isn't possible (broken measurement, or mismatched start-end time intervals) we should log errors and fall back the evaluation to the unaligned data

Signed-off-by: Ondrej Lichtner olichtne@redhat.com --- .../Perf/Measurements/BaseFlowMeasurement.py | 48 ++++++------ .../Perf/Measurements/StatCPUMeasurement.py | 14 +--- lnst/RecipeCommon/Perf/Recipe.py | 12 ++- lnst/RecipeCommon/Perf/Results.py | 74 ++++++++++++++++++- 4 files changed, 105 insertions(+), 43 deletions(-)

diff --git a/lnst/RecipeCommon/Perf/Measurements/BaseFlowMeasurement.py b/lnst/RecipeCommon/Perf/Measurements/BaseFlowMeasurement.py index dcb0d2a..a8e9328 100644 --- a/lnst/RecipeCommon/Perf/Measurements/BaseFlowMeasurement.py +++ b/lnst/RecipeCommon/Perf/Measurements/BaseFlowMeasurement.py @@ -165,40 +165,34 @@ def receiver_cpu_stats(self, value):

@property def start_timestamp(self): - return max([seq_result[0].timestamp for seq_result in self.generator_results]) + return min( + [ + self.generator_results.start_timestamp, + self.generator_cpu_stats.start_timestamp, + self.receiver_results.start_timestamp, + self.receiver_cpu_stats.start_timestamp, + ] + )

@property def end_timestamp(self): - return min([seq_result[-1].timestamp for seq_result in self.generator_results]) + return max( + [ + self.generator_results.end_timestamp, + self.generator_cpu_stats.end_timestamp, + self.receiver_results.end_timestamp, + self.receiver_cpu_stats.end_timestamp, + ] + )

- def align_data(self, start, end): + def time_slice(self, start, end): result_copy = FlowMeasurementResults(self.measurement, self.flow)

- # NOTE: iperf reports the cpu utilization for the whole test - # period, not each second, so the CPU samples cannot be aligned - result_copy.generator_cpu_stats = self.generator_cpu_stats - result_copy.receiver_cpu_stats = self.receiver_cpu_stats - - result_copy.generator_results = ParallelPerfResult() - result_copy.receiver_results = ParallelPerfResult() - - for stream in self.generator_results: - aligned_intervals = [ - interval - for interval in stream - if interval.timestamp >= start and interval.timestamp <= end - ] - - result_copy.generator_results.append(SequentialPerfResult(aligned_intervals)) - - for stream in self.receiver_results: - aligned_intervals = [ - interval - for interval in stream - if interval.timestamp >= start and interval.timestamp <= end - ] + result_copy.generator_cpu_stats = self.generator_cpu_stats.time_slice(start, end) + result_copy.receiver_cpu_stats = self.receiver_cpu_stats.time_slice(start, end)

- result_copy.receiver_results.append(SequentialPerfResult(aligned_intervals)) + result_copy.generator_results = self.generator_results.time_slice(start, end) + result_copy.receiver_results = self.generator_results.time_slice(start, end)

return result_copy

diff --git a/lnst/RecipeCommon/Perf/Measurements/StatCPUMeasurement.py b/lnst/RecipeCommon/Perf/Measurements/StatCPUMeasurement.py index 49aa262..c5b50fe 100644 --- a/lnst/RecipeCommon/Perf/Measurements/StatCPUMeasurement.py +++ b/lnst/RecipeCommon/Perf/Measurements/StatCPUMeasurement.py @@ -28,26 +28,20 @@ def utilization(self):

@property def start_timestamp(self): - return self._data["user"][0].timestamp + return min([item.start_timestamp for item in self._data.values()])

@property def end_timestamp(self): - return self._data["user"][-1].timestamp + return max([item.end_timestamp for item in self._data.values()])

- def align_data(self, start, end): + def time_slice(self, start, end): result_copy = StatCPUMeasurementResults( self.measurement, self.host, self.cpu ) for cpu_state, intervals in self._data.items(): - aligned_interval = [ - interval - for interval in intervals - if interval.timestamp >= start and interval.timestamp <= end - ] - - result_copy._data[cpu_state] = SequentialPerfResult(aligned_interval) + result_copy._data[cpu_state] = intervals.time_slice(start, end) return result_copy

diff --git a/lnst/RecipeCommon/Perf/Recipe.py b/lnst/RecipeCommon/Perf/Recipe.py index 82f9649..325e96b 100644 --- a/lnst/RecipeCommon/Perf/Recipe.py +++ b/lnst/RecipeCommon/Perf/Recipe.py @@ -3,11 +3,13 @@ from typing import Any, List, Dict

from lnst.Common.LnstError import LnstError +from lnst.Common.Logs import log_exc_traceback from lnst.Controller.Recipe import BaseRecipe from lnst.RecipeCommon.Perf.Measurements.BaseMeasurement import ( BaseMeasurement, BaseMeasurementResults, ) +from lnst.RecipeCommon.Perf.Results import EmptySlice from lnst.RecipeCommon.Perf.Results import SequentialPerfResult from lnst.RecipeCommon.Perf.Results import ParallelPerfResult

@@ -117,7 +119,7 @@ def time_aligned_results(self) -> "RecipeResults": for i, measurement_iteration in enumerate(measurement_results): aligned_measurement_results = [] for result in measurement_iteration: - aligned_measurement_result = result.align_data( + aligned_measurement_result = result.time_slice( timestamps[i][0], timestamps[i][1] ) aligned_measurement_results.append( @@ -174,7 +176,13 @@ def describe_perf_test_iteration_tweak(self, recipe_conf: RecipeConf): self.add_result(True, "\n".join(description))

def perf_report_and_evaluate(self, results: RecipeResults): - aligned_results = results.time_aligned_results + try: + aligned_results = results.time_aligned_results + except EmptySlice: + logging.error("Result time alignment impossible, falling back to unaligned results") + log_exc_traceback() + aligned_results = results +

self.perf_report(aligned_results) self.perf_evaluate(aligned_results) diff --git a/lnst/RecipeCommon/Perf/Results.py b/lnst/RecipeCommon/Perf/Results.py index 61a9ef8..aaca05f 100644 --- a/lnst/RecipeCommon/Perf/Results.py +++ b/lnst/RecipeCommon/Perf/Results.py @@ -1,6 +1,9 @@ from lnst.Common.LnstError import LnstError from lnst.Common.Utils import std_deviation

+class EmptySlice(LnstError): + pass + class PerfStatMixin(object): @property def average(self): @@ -26,6 +29,17 @@ def duration(self): def unit(self): raise NotImplementedError()

+ @property + def start_timestamp(self): + raise NotImplementedError() + + @property + def end_timestamp(self): + raise NotImplementedError() + + def time_slice(self, start, end): + raise NotImplementedError() + class PerfInterval(PerfResult): def __init__(self, value, duration, unit, timestamp): self._value = value @@ -46,9 +60,13 @@ def unit(self): return self._unit

@property - def timestamp(self): + def start_timestamp(self): return self._timestamp

+ @property + def end_timestamp(self): + return self._timestamp + self.duration + @property def std_deviation(self): return 0 @@ -57,6 +75,20 @@ def __str__(self): return "{:.2f} {} in {:.2f} seconds".format( float(self.value), self.unit, float(self.duration))

+ def time_slice(self, start, end): + if end < self.start_timestamp or start > self.end_timestamp: + raise EmptySlice( + "current start, end {} {}; request start, end {}, {}".format( + self.start_timestamp, self.end_timestamp, start, end, + ) + ) + + new_start = max(self.start_timestamp, start) + new_end = min(self.end_timestamp, end) + new_duration = new_end - new_start + new_value = self.value * (new_duration/self.duration) + return PerfInterval(new_value, new_duration, self.unit, new_start) + class PerfList(list): def __init__(self, iterable=[]): for i, item in enumerate(iterable): @@ -123,7 +155,23 @@ def __setitem__(self, i, item):

super(PerfList, self).__setitem__(i, item)

-class SequentialPerfResult(PerfResult, PerfList): + def time_slice(self, start, end): + result = self.__class__() + for item in self: + try: + item_slice = item.time_slice(start, end) + result.append(item_slice) + except EmptySlice: + continue + if len(result) == 0: + raise EmptySlice( + "current start, end {} {}; request start, end {}, {}".format( + self.start_timestamp, self.end_timestamp, start, end, + ) + ) + return result + +class SequentialPerfResult(PerfList, PerfResult): @property def value(self): return sum([i.value for i in self]) @@ -139,14 +187,24 @@ def unit(self): else: return None

-class ParallelPerfResult(PerfResult, PerfList): + @property + def start_timestamp(self): + return self[0].start_timestamp + + @property + def end_timestamp(self): + return self[-1].end_timestamp + +class ParallelPerfResult(PerfList, PerfResult): @property def value(self): return sum([i.value for i in self])

@property def duration(self): - return max([i.duration for i in self]) + min_start = min([item.start_timestamp for item in self]) + max_end = max([item.end_timestamp for item in self]) + return max_end - min_start

@property def unit(self): @@ -155,6 +213,14 @@ def unit(self): else: return None

+ @property + def start_timestamp(self): + return min([i.start_timestamp for i in self]) + + @property + def end_timestamp(self): + return max([i.end_timestamp for i in self]) + def result_averages_difference(a, b): if a is None or b is None: return None