Upload 4 files
Browse files- app.py +21 -0
- parameters.py +2 -0
- requirements.txt +0 -0
- utils.py +214 -0
app.py
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import utils
|
2 |
+
import gradio as gr
|
3 |
+
import parameters
|
4 |
+
|
5 |
+
with gr.Blocks() as demo:
|
6 |
+
# gr.Tab(label="Error Metrics Dashboard")
|
7 |
+
gr.Markdown("# Error Metrics Dashboard")
|
8 |
+
with gr.Row():
|
9 |
+
project = gr.Textbox("Project Name", value= parameters.project)
|
10 |
+
start_date = gr.Textbox(label="Start Date (YYYY-MM-DD)", value="2025-02-20")
|
11 |
+
end_date = gr.Textbox(label="End Date (YYYY-MM-DD)", value="2025-02-21")
|
12 |
+
submit_btn = gr.Button("Process")
|
13 |
+
output = gr.Textbox(label="Results", lines=10)
|
14 |
+
|
15 |
+
submit_btn.click(
|
16 |
+
fn=utils.process_dates,
|
17 |
+
inputs=[start_date, end_date,project],
|
18 |
+
outputs=output
|
19 |
+
)
|
20 |
+
|
21 |
+
demo.launch(share = True)
|
parameters.py
ADDED
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
1 |
+
api_key = '9tcxDNK6Wtfmf5C5ogWpqK7tr'
|
2 |
+
project ='vf-ai-prod'
|
requirements.txt
ADDED
Binary file (3.37 kB). View file
|
|
utils.py
ADDED
@@ -0,0 +1,214 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import datetime
|
2 |
+
import json
|
3 |
+
import os
|
4 |
+
from opik import Opik
|
5 |
+
import parameters
|
6 |
+
from collections import defaultdict
|
7 |
+
from concurrent.futures import ThreadPoolExecutor, as_completed
|
8 |
+
|
9 |
+
|
10 |
+
class DateTimeEncoder(json.JSONEncoder):
|
11 |
+
def default(self, obj):
|
12 |
+
if isinstance(obj, datetime.datetime):
|
13 |
+
return obj.isoformat()
|
14 |
+
return super().default(obj)
|
15 |
+
|
16 |
+
def get_trace_content(opik, trace_id):
|
17 |
+
try:
|
18 |
+
trace_content = opik.get_trace_content(trace_id)
|
19 |
+
return trace_content.dict()
|
20 |
+
except Exception as e:
|
21 |
+
print(f"Error getting trace content {trace_id}: {e}")
|
22 |
+
return None
|
23 |
+
|
24 |
+
def get_span_content(opik, trace_id, span):
|
25 |
+
try:
|
26 |
+
content = opik.get_span_content(span.id)
|
27 |
+
return {"trace_id": trace_id, "span_id": span.id, "content": content.dict()}
|
28 |
+
except Exception as e:
|
29 |
+
print(f"Error getting span content {span.id}: {e}")
|
30 |
+
return None
|
31 |
+
|
32 |
+
def get_traces_on_date(start_date_str, end_date_str, project_name, api_key,max_workers=10):
|
33 |
+
try:
|
34 |
+
print("Step 1: Converting date strings")
|
35 |
+
date = datetime.date.fromisoformat(start_date_str)
|
36 |
+
start_date_str = date.isoformat() + "T00:00:00Z"
|
37 |
+
|
38 |
+
if not end_date_str:
|
39 |
+
end_date = date + datetime.timedelta(days=1)
|
40 |
+
end_date_str = end_date.isoformat() + "T00:00:00Z"
|
41 |
+
else:
|
42 |
+
end_date = datetime.date.fromisoformat(end_date_str)
|
43 |
+
end_date_str = end_date.isoformat() + "T00:00:00Z"
|
44 |
+
|
45 |
+
print(f"Start: {start_date_str} and end: {end_date_str}")
|
46 |
+
filter_string = f'start_time >= "{start_date_str}" and end_time <= "{end_date_str}"'
|
47 |
+
print("Filter string: ", filter_string)
|
48 |
+
|
49 |
+
print("Step 2: Initializing Opik client")
|
50 |
+
try:
|
51 |
+
opik = Opik(api_key=api_key, project_name=project_name, workspace='verba-tech-ninja')
|
52 |
+
print("Opik client initialized successfully")
|
53 |
+
except Exception as e:
|
54 |
+
print(f"Error initializing Opik client: {e}")
|
55 |
+
return [], []
|
56 |
+
|
57 |
+
print("Step 3: Searching traces")
|
58 |
+
try:
|
59 |
+
traces = opik.search_traces(filter_string=filter_string, project_name=project_name)
|
60 |
+
print("Total searches: ", len(traces))
|
61 |
+
except Exception as e:
|
62 |
+
print(f"Error searching traces: {e}")
|
63 |
+
return [], []
|
64 |
+
|
65 |
+
print("Step 4: Processing traces in parallel")
|
66 |
+
all_traces_content = []
|
67 |
+
try:
|
68 |
+
with ThreadPoolExecutor(max_workers=max_workers) as executor:
|
69 |
+
future_to_trace = {executor.submit(get_trace_content, opik, trace.id): trace for trace in traces}
|
70 |
+
for future in as_completed(future_to_trace):
|
71 |
+
result = future.result()
|
72 |
+
if result:
|
73 |
+
all_traces_content.append(result)
|
74 |
+
print(f"Completed processing {len(all_traces_content)} traces")
|
75 |
+
except Exception as e:
|
76 |
+
print(f"Error processing traces in parallel: {e}")
|
77 |
+
|
78 |
+
print("Step 5: Processing spans in parallel")
|
79 |
+
all_spans_content = []
|
80 |
+
try:
|
81 |
+
with ThreadPoolExecutor(max_workers=max_workers) as executor:
|
82 |
+
future_to_span = {}
|
83 |
+
for i, trace in enumerate(traces):
|
84 |
+
try:
|
85 |
+
print(f"Searching spans for trace_id: {trace.id}:{i+1}/{len(traces)}")
|
86 |
+
spans = opik.search_spans(project_name=parameters.project, trace_id=trace.id)
|
87 |
+
print(f"Found {len(spans)} spans for trace_id: {trace.id}")
|
88 |
+
for span in spans:
|
89 |
+
future_to_span[executor.submit(get_span_content, opik, trace.id, span)] = span
|
90 |
+
except Exception as e:
|
91 |
+
print(f"Error searching spans for trace {trace.id}: {e}")
|
92 |
+
|
93 |
+
for future in as_completed(future_to_span):
|
94 |
+
result = future.result()
|
95 |
+
if result:
|
96 |
+
all_spans_content.append(result)
|
97 |
+
print(f"Completed processing {len(all_spans_content)} spans")
|
98 |
+
except Exception as e:
|
99 |
+
print(f"Error processing spans in parallel: {e}")
|
100 |
+
|
101 |
+
print("Step 6: Saving to JSON files")
|
102 |
+
traces_file = 'all_traces_content.json'
|
103 |
+
spans_file = 'all_spans_content.json'
|
104 |
+
try:
|
105 |
+
if os.path.exists(traces_file):
|
106 |
+
os.remove(traces_file)
|
107 |
+
print(f"Removed existing {traces_file}")
|
108 |
+
if os.path.exists(spans_file):
|
109 |
+
os.remove(spans_file)
|
110 |
+
print(f"Removed existing {spans_file}")
|
111 |
+
|
112 |
+
print(f"Writing {len(all_traces_content)} traces to {traces_file}")
|
113 |
+
with open(traces_file, 'w') as f:
|
114 |
+
json.dump(all_traces_content, f, indent=2, cls=DateTimeEncoder)
|
115 |
+
print(f"Saved traces to {traces_file}")
|
116 |
+
|
117 |
+
print(f"Writing {len(all_spans_content)} spans to {spans_file}")
|
118 |
+
with open(spans_file, 'w') as f:
|
119 |
+
json.dump(all_spans_content, f, indent=2, cls=DateTimeEncoder)
|
120 |
+
print(f"Saved spans to {spans_file}")
|
121 |
+
except Exception as e:
|
122 |
+
print(f"Error saving to JSON files: {e}")
|
123 |
+
with open('partial_traces_content.json', 'w') as f:
|
124 |
+
json.dump(all_traces_content, f, indent=2, cls=DateTimeEncoder)
|
125 |
+
with open('partial_spans_content.json', 'w') as f:
|
126 |
+
json.dump(all_spans_content, f, indent=2, cls=DateTimeEncoder)
|
127 |
+
print("Saved partial data to partial_traces_content.json and partial_spans_content.json")
|
128 |
+
|
129 |
+
print("Step 7: Returning results")
|
130 |
+
return all_traces_content, all_spans_content
|
131 |
+
|
132 |
+
except Exception as e:
|
133 |
+
print(f"Main function error: {e}")
|
134 |
+
return [], []
|
135 |
+
|
136 |
+
def find_errors_and_metrics(traces, spans):
|
137 |
+
try:
|
138 |
+
print("Step 8: Analyzing outputs for errors")
|
139 |
+
error_spans = []
|
140 |
+
error_metrics = defaultdict(list)
|
141 |
+
|
142 |
+
|
143 |
+
for span in spans:
|
144 |
+
content = span['content']
|
145 |
+
output = content.get("output")
|
146 |
+
if isinstance(output, dict) and 'output' in output:
|
147 |
+
output_value = output.get("output")
|
148 |
+
else:
|
149 |
+
output_value = output
|
150 |
+
# print(f"Span output_value: {output_value}")
|
151 |
+
if (output_value is None or (isinstance(output, list) and len(output) ==0)):
|
152 |
+
# print(output_value)
|
153 |
+
error_spans.append({
|
154 |
+
"trace_id": span["trace_id"],
|
155 |
+
"span_id": span["span_id"],
|
156 |
+
"error_content": output
|
157 |
+
})
|
158 |
+
error_metrics["empty_output"].append(span["trace_id"])
|
159 |
+
|
160 |
+
print(f"Found {len(error_spans)} outputs with errors (empty/null)")
|
161 |
+
|
162 |
+
print("Step 9: Saving error spans")
|
163 |
+
error_file = 'error_spans.json'
|
164 |
+
try:
|
165 |
+
if os.path.exists(error_file):
|
166 |
+
os.remove(error_file)
|
167 |
+
print(f"Removed existing {error_file}")
|
168 |
+
print(f"Writing {len(error_spans)} error outputs to {error_file}")
|
169 |
+
with open(error_file, 'w') as f:
|
170 |
+
json.dump(error_spans, f, indent=2, cls=DateTimeEncoder)
|
171 |
+
print(f"Saved error outputs to {error_file}")
|
172 |
+
except Exception as e:
|
173 |
+
print(f"Error saving error spans: {e}")
|
174 |
+
|
175 |
+
print("Step 10: Calculating metrics")
|
176 |
+
metrics = {
|
177 |
+
"total_errors": len(error_spans),
|
178 |
+
"unique_error_types": {
|
179 |
+
error_type: {
|
180 |
+
"count": len(set(trace_ids)),
|
181 |
+
"trace_ids": list(set(trace_ids))
|
182 |
+
}
|
183 |
+
for error_type, trace_ids in error_metrics.items()
|
184 |
+
}
|
185 |
+
}
|
186 |
+
print(f"Metrics calculated: {len(metrics['unique_error_types'])} unique error types")
|
187 |
+
return metrics
|
188 |
+
|
189 |
+
except Exception as e:
|
190 |
+
print(f"Error in find_errors_and_metrics: {e}")
|
191 |
+
return {}
|
192 |
+
|
193 |
+
def process_dates(start_date, end_date,project):
|
194 |
+
try:
|
195 |
+
print("Pipeline Start: Processing dates")
|
196 |
+
traces, spans = get_traces_on_date(start_date, end_date, project, parameters.api_key)
|
197 |
+
metrics = find_errors_and_metrics(traces, spans)
|
198 |
+
|
199 |
+
output = f"Total Empty/Null Outputs Found: {metrics.get('total_errors', 0)}\n\n"
|
200 |
+
output += f"Total Traces Found: {len(traces)}\n"
|
201 |
+
output += f"Total Spans Processed: {len(spans)}\n\n"
|
202 |
+
output += "Error Metrics:\n"
|
203 |
+
if metrics.get('unique_error_types', {}):
|
204 |
+
for error_type, data in metrics.get('unique_error_types', {}).items():
|
205 |
+
output += f"{error_type.replace('_', ' ').title()}: {data['count']} unique occurrences\n"
|
206 |
+
output += f"Trace IDs: {', '.join(data['trace_ids'][:5])}{'...' if len(data['trace_ids']) > 5 else ''}\n\n"
|
207 |
+
else:
|
208 |
+
output += "No empty or null outputs detected.\n"
|
209 |
+
|
210 |
+
print("Pipeline End: Results formatted")
|
211 |
+
return output
|
212 |
+
except Exception as e:
|
213 |
+
print(f"Error processing dates: {e}")
|
214 |
+
return f"Error processing dates: {e}"
|