Skip to content

Commit

Permalink
can plot churn stuffgit add .git add .!
Browse files Browse the repository at this point in the history
  • Loading branch information
@deryacog
deryacog committed Jan 10, 2025
1 parent 151397e commit 9ba6dd1
Show file tree
Hide file tree
Showing 5 changed files with 71 additions and 103 deletions.
93 changes: 11 additions & 82 deletions mergetb/average_churn_data.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,102 +5,31 @@
from average_data import metrics_to_extract, get_data, calculate_average_data
import extract_raw_data

from extract_raw_churn_data import metrics_interval

if __name__ == "__main__":
if len(sys.argv) < 3:
if len(sys.argv) < 1:
print("Usage: python average_data.py <directory> <duration>")
sys.exit(1)

directory = sys.argv[1]
duration = float(sys.argv[2])
data = get_data(directory)

average_data = {}

print(f"variables: {data.keys()}")

for variable_name, variable_data in data.items():
# for variable_name, variable_data in ["duplicates"]:
# for variable_name, variable_data in ["duplicates"]:
# variable_name = "retry"
# variable_data = data[variable_name]

average_data[variable_name] = {}
print(f"runs: {variable_data.keys()}")
for run, runs in variable_data.items():
# run = "4"
average_data[variable_name][run] = {}
print("Getting data for: ", variable_name, run)

for time_index, time_data in variable_data[run].items():
# for time_index, time_data in runs.items():
if time_index != "0":
prev_step = variable_data[run][str(int(time_index) - metrics_interval)]
# print(f"time_index: {time_index}, prev_step: {prev_step}")
# print(f"prev_step data: {prev_step['loopix_incoming_messages'], prev_step['loopix_number_of_proxy_requests']}")
# print(f"time_data data: {time_data['loopix_incoming_messages'], time_data['loopix_number_of_proxy_requests']}")

for metric in extract_raw_data.metrics_to_extract:
if metric in ["loopix_bandwidth_bytes", "loopix_incoming_messages", "loopix_number_of_proxy_requests"]:
# print(f"{time_index}--------------------------------")
prev_step_data = np.array(prev_step[metric])
time_data_data = np.array(time_data[metric])
# print(f"prev_step_data: {prev_step_data}")
# print(f"time_data_data: {time_data_data}")
# print(f"data[{metric}]:")
if len(prev_step_data) < len(time_data_data):
data[metric] = time_data_data[:len(prev_step_data)] - prev_step_data
elif len(prev_step_data) > len(time_data_data):
data[metric] = time_data_data - prev_step_data[:len(time_data_data)]
else:
data[metric] = time_data_data - prev_step_data
# print(f"{data[metric]}")

elif metric == "loopix_start_time_seconds":
data[metric] = time_data[metric]
else:
# print("--------------------------------")
prev_step_sum = np.array(prev_step[metric]["sum"])
time_data_sum = np.array(time_data[metric]["sum"])
# print(f"prev_step_sum: {prev_step_sum}")
# print(f"time_data_sum: {time_data_sum}")

if len(prev_step_sum) < len(time_data_sum):
data[metric]["sum"] = time_data_sum[:len(prev_step_sum)] - prev_step_sum
elif len(prev_step_sum) > len(time_data_sum):
data[metric]["sum"] = time_data_sum - prev_step_sum[:len(time_data_sum)]
else:
data[metric]["sum"] = time_data_sum - prev_step_sum

prev_step_count = np.array(prev_step[metric]["count"])
time_data_count = np.array(time_data[metric]["count"])
# print(f"prev_step_count: {prev_step_count}")
# print(f"time_data_count: {time_data_count}")

if len(prev_step_count) < len(time_data_count):
data[metric]["count"] = time_data_count[:len(prev_step_count)] - prev_step_count
elif len(prev_step_count) > len(time_data_count):
data[metric]["count"] = time_data_count - prev_step_count[:len(time_data_count)]
else:
data[metric]["count"] = time_data_count - prev_step_count

# print(f"data[{metric}]: {data[metric]}")


else:
data = time_data
print(f"time_index: {time_index}, data: {data['loopix_incoming_messages'], data['loopix_number_of_proxy_requests']}")
for variable, runs in data.items():
average_data[variable] = {}
for try_index, kill_runs in runs.items():
average_data[variable][try_index] = {}
for kill, runs in kill_runs.items():
print(runs.keys())
print(f"try_index: {try_index}")
print(f"run: {kill}")


average_data[variable_name][run][time_index] = {}
print(f"data: {data}")
# data = time_data
average_data[variable_name][run][time_index] = calculate_average_data(data, metrics_interval)
average_data[variable][try_index][kill] = calculate_average_data(runs, duration)

with open(os.path.join(directory, 'average_data.json'), 'w') as f:
json.dump(average_data, f, indent=2)




2 changes: 1 addition & 1 deletion mergetb/average_data.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,7 @@ def calculate_average_data(data, duration):
return results

if __name__ == "__main__":
if len(sys.argv) < 1:
if len(sys.argv) < 3:
print("Usage: python average_data.py <directory> <duration>")
sys.exit(1)

Expand Down
6 changes: 2 additions & 4 deletions mergetb/extract_raw_churn_data.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ def get_metrics_data(data_dir, path_length, n_clients, results, variable, index)
create_results_dict(results, metrics_to_extract)

directory = os.path.join(data_dir, variable)
metrics_file = os.path.join(directory, f"metrics_4.txt")
metrics_file = os.path.join(directory, f"metrics.txt")

print(f"Getting metrics data from {metrics_file}")
if os.path.exists(metrics_file):
Expand Down Expand Up @@ -96,9 +96,7 @@ def main():

node_dir = os.path.join(run_dir, f"{i}_{j}")
print(node_dir)

metrics_file_name = f"metrics_{4}.txt"



nodes = os.listdir(node_dir)
for enum_node, node in enumerate(nodes):
Expand Down
69 changes: 55 additions & 14 deletions mergetb/plot_data_churn.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,8 +3,12 @@
import json
import sys
import os
from plot_data import get_data, plot_latency_components, plot_reliability, plot_incoming_messages, plot_latency, plot_bandwidth, plot_latency_and_bandwidth, plot_reliability_latency
from plot_data import get_data, save_plot_directory, plot_latency_components, plot_reliability, plot_incoming_messages, plot_latency, plot_bandwidth, plot_latency_and_bandwidth, plot_reliability_latency
from matplotlib.ticker import FuncFormatter

def log_format(y, _):
return f"{y:.0f}"

x_axis_name = {"lambda_loop": "Loop and Drop Messages per Second", "lambda_payload": "Payload Messages per Second from the Client"}

if __name__ == "__main__":
Expand All @@ -18,16 +22,53 @@
duration = int(sys.argv[3])
data = get_data(directory)

for variable, run in data.items():
print(f"Plotting {variable}")
for run_index, run_data in run.items():
plot_dir = os.path.join(directory, "plots", variable, run_index)
print(f"Plotting {run_index} with plot_dir {plot_dir}")
plot_latency_components(plot_dir, path_length, variable, run_data)
plot_reliability(plot_dir, variable, run_data)
plot_incoming_messages(plot_dir, variable, run_data)
plot_latency(plot_dir, variable, run_data)
plot_bandwidth(plot_dir, duration, variable, run_data)
plot_latency_and_bandwidth(plot_dir, variable, run_data)
# plot_reliability_incoming_latency(directory, variable, run)
plot_reliability_latency(plot_dir, variable, run_data)
retry_levels = data["retry"].keys()
points_in_time = list(data["retry"]["0"].keys())
points_in_time_labels = [
"All Nodes Live", "1 Node Down", "2 Nodes Down", "3 Nodes Down"
]

reliability_data = {
retry: [
data["retry"][retry][str(pt)]["loopix_reliability"]
for pt in points_in_time
]
for retry in retry_levels
}

plt.figure(figsize=(12, 8))

bar_width = 0.2
x_indexes = range(len(points_in_time_labels))
offset = 0


plt.yscale('log')
plt.gca().yaxis.set_major_formatter(FuncFormatter(log_format))

for retry, reliabilities in reliability_data.items():
reliability = [r*100 for r in reliabilities]
plt.bar(
[x + offset for x in x_indexes],
reliability,
width=bar_width,
label=f"{retry} Retry",
)
offset += bar_width

# Customizing the plot
plt.title("Reliability vs Node Status", fontsize=16)
plt.xlabel("Node Status", fontsize=14)
plt.ylabel("Reliability", fontsize=14)
plt.xticks([x + bar_width * (len(retry_levels) - 1) / 2 for x in x_indexes], points_in_time_labels, fontsize=12)
plt.legend(fontsize=12)
plt.grid(axis='y', linestyle='--', alpha=0.7)

# Save and close the plot
save_plot_directory(directory)
plt.savefig(os.path.join(directory, "reliability_churn_bar_chart.png"))
plt.clf()
plt.close()



4 changes: 2 additions & 2 deletions mergetb/run_churn_data_processing.bash
View file
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,6 @@ source ../venv/bin/activate

python3 extract_raw_churn_data.py "$DATA_DIR" "$PATH_LENGTH" "$N_CLIENTS"

python3 average_churn_data.py "$DATA_DIR" "$DURATION"
# python3 average_churn_data.py "$DATA_DIR" "$DURATION"

python3 plot_data_churn.py "$DATA_DIR" "$PATH_LENGTH" "$DURATION"
# python3 plot_data_churn.py "$DATA_DIR" "$PATH_LENGTH" "$DURATION"

0 comments on commit 9ba6dd1

Please sign in to comment.