-
Notifications
You must be signed in to change notification settings - Fork 0
/
main.py
279 lines (248 loc) · 10 KB
/
main.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
# app.py
import uvicorn
from fastapi import FastAPI, Request, Form
from fastapi.responses import HTMLResponse
from typing import List
import yaml
import subprocess
import statistics
from math import isnan, fsum
from tabulate import tabulate
import os
app = FastAPI()
# Define ANSI colors for different modes (simplified for web)
COLORS = {
'dark': {
'highlight': 'green',
'header': 'white',
'zone': 'skyblue',
'dim': 'grey',
'gray': 'lightgrey',
'teal': 'teal',
'orange': 'orange'
}
}
TRANSFER_RATES = {
'internet': 0.09, # $/GB to internet
'within_region': 0.01, # $/GB within region
'cross_region': 0.02 # $/GB across regions
}
STORAGE_COSTS = {'standard': 0.023, 'glacier': 0.004} # $/GB/month
def harmonic_mean(data):
"""Calculate the harmonic mean of a list of numbers."""
if not data or any(d == 0 for d in data):
return float('nan') # Avoid division by zero
return len(data) / fsum(1 / d for d in data)
def stability_metric(prices):
"""Calculate the volatility as the difference between the max and min prices."""
if not prices:
return float('nan')
return max(prices) - min(prices)
def calculate_vcpu_mins(x_coverage, vcpu_min_per_x):
"""Calculate total vCPU-minutes based on coverage."""
return x_coverage * vcpu_min_per_x
def calculate_storage_costs(size_gb, storage_type):
"""Calculate monthly storage costs for a given size."""
return size_gb * STORAGE_COSTS[storage_type]
def calculate_transfer_costs(size_gb, transfer_type):
"""Calculate transfer costs based on the type."""
return size_gb * TRANSFER_RATES[transfer_type]
def get_spot_price(instance_type, zone, profile=None):
region = zone[:-1]
try:
cmd = [
"aws", "ec2", "describe-spot-price-history",
"--instance-types", instance_type,
"--availability-zone", zone,
"--product-description", "Linux/UNIX",
"--query", "SpotPriceHistory[0].SpotPrice",
"--output", "text", "--region", region,
]
if profile:
cmd.extend(["--profile", profile])
result = subprocess.check_output(cmd).decode().strip()
return float(result)
except subprocess.CalledProcessError as e:
print(f"Failed to fetch price for {zone}: {e}")
return float('nan')
def collect_spot_prices(instance_types, zones, profile=None):
"""Fetch spot prices for each instance type in the given zones."""
spot_data = {}
for instance_type in instance_types:
spot_data[instance_type] = {}
for zone in zones:
price = get_spot_price(instance_type, zone, profile)
spot_data[instance_type][zone] = price if not isnan(price) else float('nan')
return spot_data
def extract_instances(config, partition_name):
"""Extract instance types from the given partition in the YAML configuration."""
instances = []
for queue in config.get('Scheduling', {}).get('SlurmQueues', []):
if queue.get('Name') == partition_name:
for resource in queue.get('ComputeResources', []):
for instance in resource.get('Instances', []):
instances.append(instance.get('InstanceType'))
if not instances:
raise ValueError(f"No instances found for partition: {partition_name}")
return instances
def get_best_worst(stats, attr):
"""Extract the best (min) and worst (max) values for a specific numeric attribute."""
values = [getattr(z, attr) for z in stats if not isnan(getattr(z, attr))]
if not values:
return float('nan'), float('nan') # Handle case where all values are NaN
return min(values), max(values)
class ZoneStat:
def __init__(self, zone_name):
self.zone_name = zone_name
self.instances = 0
self.prices = []
self.avg_price = float('nan')
self.min_price = float('nan')
self.max_price = float('nan')
self.harmonic_price = float('nan')
self.stability = float('nan')
self.cost_per_vcpu = float('nan')
self.est_cost = float('nan')
self.bam_size = float('nan')
self.cram_size = float('nan')
self.vcf_size = float('nan')
self.gvcf_size = float('nan')
self.bcf_size = float('nan')
self.gbcf_size = float('nan')
self.fastq_size = float('nan')
self.qc_size = float('nan') # Added for QC data size
def calculate_statistics(self, spot_data, vcpu_mins, args):
# Calculate the various statistics for this zone
self.prices = [spot_data[instance].get(self.zone_name, float('nan')) for instance in spot_data]
self.prices = [p for p in self.prices if not isnan(p)]
if self.prices:
self.instances = len(self.prices)
self.avg_price = statistics.mean(self.prices)
self.min_price = min(self.prices)
self.max_price = max(self.prices)
self.harmonic_price = harmonic_mean(self.prices)
self.stability = stability_metric(self.prices)
self.cost_per_vcpu = self.avg_price
self.est_cost = (vcpu_mins / 60) * self.avg_price
# Use args to calculate sizes
self.bam_size = args['x_coverage'] * args['bam_size_per_x'] if args['bam_size_per_x'] else 0
self.cram_size = args['x_coverage'] * args['cram_size_per_x'] if args['cram_size_per_x'] else 0
self.vcf_size = args['x_coverage'] * args['vcf_size_per_x'] if args['vcf_size_per_x'] else 0
self.gvcf_size = args['x_coverage'] * args['gvcf_size_per_x'] if args['gvcf_size_per_x'] else 0
self.bcf_size = args['x_coverage'] * args['bcf_size_per_x'] if args['bcf_size_per_x'] else 0
self.gbcf_size = args['x_coverage'] * args['gbcf_size_per_x'] if args['gbcf_size_per_x'] else 0
self.fastq_size = args['x_coverage'] * args['input_data_size_per_x'] if args['input_data_size_per_x'] else 0
self.qc_size = args['x_coverage'] * args['qc_size_per_x'] if args['qc_size_per_x'] else 0 # QC data size
else:
# No prices, leave values as NaN or zero
pass
@app.get("/", response_class=HTMLResponse)
async def index():
html_content = """
<html>
<head>
<title>Genomic Workflow Cost Calculator</title>
</head>
<body>
<h1>Genomic Workflow Cost Calculator</h1>
<form action="/calculate" method="post">
<label>Availability Zones (comma-separated):</label><br>
<input type="text" name="zones" value="us-west-2a,us-west-2b,us-west-2c"><br><br>
<label>Coverage (X):</label><br>
<input type="number" name="x_coverage" value="30.1" step="0.1"><br><br>
<label>vCPU-minutes per X coverage:</label><br>
<input type="number" name="vcpu_min_per_x" value="3.7" step="0.1"><br><br>
<!-- Add more input fields as needed -->
<input type="submit" value="Calculate">
</form>
</body>
</html>
"""
return html_content
@app.post("/calculate", response_class=HTMLResponse)
async def calculate(request: Request,
zones: str = Form(...),
x_coverage: float = Form(...),
vcpu_min_per_x: float = Form(...)):
# Default arguments (you can expose these in the form if needed)
args = {
'x_coverage': x_coverage,
'vcpu_min_per_x': vcpu_min_per_x,
'bam_size_per_x': 1.0,
'cram_size_per_x': 1.0,
'vcf_size_per_x': 1.0,
'gvcf_size_per_x': 1.0,
'bcf_size_per_x': 1.0,
'gbcf_size_per_x': 1.0,
'qc_size_per_x': 0.5,
'input_data_size_per_x': 1.0,
'mode': 'dark',
'partition': 'i192',
'profile': None,
'input': 'config/day_cluster/prod_cluster.yaml'
}
# Load YAML configuration
with open(args['input'], 'r') as f:
config = yaml.safe_load(f)
# Extract instances from the specified partition
try:
instance_types = extract_instances(config, args['partition'])
except ValueError as e:
return f"<h3>Error: {e}</h3>"
zone_list = [zone.strip() for zone in zones.split(',')]
spot_data = collect_spot_prices(instance_types, zone_list, args['profile'])
vcpu_mins = calculate_vcpu_mins(args['x_coverage'], args['vcpu_min_per_x'])
# Calculate statistics for each zone
zone_stats = []
for zone in zone_list:
zone_stat = ZoneStat(zone)
zone_stat.calculate_statistics(spot_data, vcpu_mins, args)
zone_stats.append(zone_stat)
# Generate table data
headers = [
"Availability Zone", "Instances", "Avg Spot Price", "Min Spot Price",
"Max Spot Price", "Harmonic Mean Price", "Stability (Max-Min)",
"BAM (GB)", "CRAM (GB)", "VCF (GB)", "gVCF (GB)", "BCF (GB)", "gBCF (GB)",
"Cost per vCPU", "Est. EC2 Cost"
]
table = []
for index, z in enumerate(zone_stats, 1):
row = [
f"{index}. {z.zone_name}",
z.instances,
f"{z.avg_price:.4f}",
f"{z.min_price:.4f}",
f"{z.max_price:.4f}",
f"{z.harmonic_price:.4f}",
f"{z.stability:.4f}",
f"{z.bam_size:.4f}",
f"{z.cram_size:.4f}",
f"{z.vcf_size:.4f}",
f"{z.gvcf_size:.4f}",
f"{z.bcf_size:.4f}",
f"{z.gbcf_size:.4f}",
f"{z.cost_per_vcpu:.4f}",
f"{z.est_cost:.4f}"
]
table.append(row)
# Convert table to HTML
table_html = tabulate(table, headers=headers, tablefmt="html")
# Generate the response HTML
html_content = f"""
<html>
<head>
<title>Calculation Results</title>
</head>
<body>
<h1>Calculation Results</h1>
<p>{args['x_coverage']}-cov genome @ {args['vcpu_min_per_x']} vCPU-min per coverage</p>
{table_html}
<br>
<a href="/">Go Back</a>
</body>
</html>
"""
return html_content
if __name__ == "__main__":
# Run the app with: uvicorn app:app --reload
uvicorn.run(app, host="0.0.0.0", port=8000)