Skip to content

Remove ROSA cluster discovery from fetch_pricing #17

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jeremyeder merged 1 commit into from
May 20, 2026
+72 −205
Merged

Remove ROSA cluster discovery from fetch_pricing #17

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 0 additions & 19 deletions mcp_server.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -954,25 +954,6 @@ def cloud_pricing_lookup(
return json.dumps({"pricing": _rows_to_dicts(rows), "count": len(rows)}, default=str)


@mcp.tool(
annotations={"readOnlyHint": True, "openWorldHint": False},
)
def rosa_cluster_costs() -> str:
"""Get estimated monthly costs for OpenShift/ROSA cluster nodes.

Joins instance types with EC2 pricing to estimate costs.
Requires pricing.db with both AWS pricing and ROSA cluster discovery data.
"""
conn = _get_conn()
if not _db_attached(conn, "pricing"):
return json.dumps({"error": "pricing.db not attached. Run: uv run scripts/fetch_pricing.py"})

rows = conn.execute("SELECT * FROM pricing.v_rosa_estimated_cost ORDER BY estimated_monthly_cost DESC").fetchall()
if not rows:
return json.dumps({"error": "No ROSA cluster data. Run: uv run scripts/fetch_pricing.py (requires oc access)"})
return json.dumps({"clusters": _rows_to_dicts(rows), "count": len(rows)}, default=str)


# ---------------------------------------------------------------------------
# GitHub tools
# ---------------------------------------------------------------------------
Expand Down
256 changes: 71 additions & 185 deletions scripts/fetch_pricing.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,6 @@

AWS: fetches from public pricing bulk JSON files (no auth needed).
Claude: uses published Anthropic per-token pricing (hardcoded, no API needed).
ROSA: uses `oc get nodes` via subprocess (optional, skipped if oc unavailable).

A discount factor (e.g., 0.85 for 15% off) can be stored via --discount.
All list prices are stored at face value; discount is applied in views.
Expand All @@ -21,16 +20,13 @@
uv run scripts/fetch_pricing.py # fetch all
uv run scripts/fetch_pricing.py --aws-only # AWS pricing only
uv run scripts/fetch_pricing.py --claude-only # Claude model pricing only
uv run scripts/fetch_pricing.py --rosa-only # ROSA cluster discovery only
uv run scripts/fetch_pricing.py --skip-rosa # skip ROSA (no oc needed)
uv run scripts/fetch_pricing.py --regions us-east-1 # single region
uv run scripts/fetch_pricing.py --discount 0.85 # set 15% discount factor
"""

import argparse
import json
import sqlite3
import subprocess
import sys
import tempfile
from datetime import datetime, timezone
Expand Down Expand Up @@ -74,6 +70,25 @@
{"model": "claude-3-haiku", "input_per_mtok": 0.25, "output_per_mtok": 1.25},
]

# ROSA managed service fees (on top of EC2 infrastructure costs)
# Source: https://aws.amazon.com/rosa/pricing/
ROSA_SERVICE_FEES = [
{
"sku": "rosa-worker-fee",
"description": "ROSA managed service fee per worker node per hour",
"usage_type": "ROSA-Worker",
"unit": "Hrs",
"price_per_unit": 0.171,
},
{
"sku": "rosa-hcp-cluster-fee",
"description": "ROSA HCP cluster management fee per cluster per hour",
"usage_type": "ROSA-HCP-Cluster",
"unit": "Hrs",
"price_per_unit": 0.25,
},
]


# ---------------------------------------------------------------------------
# Schema
Expand Down Expand Up @@ -104,17 +119,6 @@
UNIQUE(provider, service, region, sku, usage_type, unit, tier_start)
);

CREATE TABLE IF NOT EXISTS rosa_cluster_instance (
cluster_name TEXT NOT NULL,
environment TEXT,
node_name TEXT NOT NULL,
instance_type TEXT NOT NULL,
role TEXT,
availability_zone TEXT,
fetched_at TEXT NOT NULL,
PRIMARY KEY (cluster_name, node_name)
);

CREATE TABLE IF NOT EXISTS _meta (
key TEXT PRIMARY KEY,
value TEXT
Expand All @@ -125,8 +129,6 @@
CREATE INDEX IF NOT EXISTS idx_pricing_instance_type ON cloud_pricing(instance_type);
CREATE INDEX IF NOT EXISTS idx_pricing_region ON cloud_pricing(region);
CREATE INDEX IF NOT EXISTS idx_pricing_model ON cloud_pricing(model_name);
CREATE INDEX IF NOT EXISTS idx_rosa_instance_type ON rosa_cluster_instance(instance_type);

-- Views (discount_factor from _meta, defaults to 1.0)
CREATE VIEW IF NOT EXISTS v_ec2_ondemand AS
SELECT instance_type, instance_family, vcpu, memory_gb, region,
Expand All @@ -141,20 +143,6 @@
WHERE provider = 'aws' AND service = 'ec2' AND usage_type = 'OnDemand'
ORDER BY instance_family, vcpu;

CREATE VIEW IF NOT EXISTS v_rosa_estimated_cost AS
SELECT r.cluster_name, r.environment, r.instance_type, r.role,
COUNT(*) AS node_count,
p.price_per_unit AS list_hourly_per_node,
ROUND(COUNT(*) * p.price_per_unit * COALESCE(
(SELECT CAST(value AS REAL) FROM _meta WHERE key = 'discount_factor'), 1.0
) * 730, 2) AS estimated_monthly_cost
FROM rosa_cluster_instance r
LEFT JOIN cloud_pricing p ON r.instance_type = p.instance_type
AND p.provider = 'aws' AND p.service = 'ec2'
AND p.usage_type = 'OnDemand'
AND p.region = SUBSTR(r.availability_zone, 1, LENGTH(r.availability_zone) - 1)
GROUP BY r.cluster_name, r.environment, r.instance_type, r.role;

CREATE VIEW IF NOT EXISTS v_vertex_ai_pricing AS
SELECT model_name, description, usage_type, unit, price_per_unit,
tier_start, tier_end
Expand Down Expand Up @@ -239,7 +227,9 @@ def _extract_ondemand_prices(terms: dict, sku: str) -> list[dict]:
"""Extract OnDemand price dimensions for a given SKU."""
results = []
on_demand = terms.get("OnDemand", {}).get(sku, {})
for _term_key, term_details in on_demand.items() if isinstance(on_demand, dict) else []:
for _term_key, term_details in (
on_demand.items() if isinstance(on_demand, dict) else []
):
for _dim_key, dim in term_details.get("priceDimensions", {}).items():
price_str = dim.get("pricePerUnit", {}).get("USD", "0")
try:
Expand Down Expand Up @@ -284,7 +274,9 @@ def fetch_aws_pricing(regions: list[str], now: str) -> list[dict]:
try:
# Download to temp file to handle large files (EC2 ~150MB/region)
with tempfile.NamedTemporaryFile(suffix=".json", delete=True) as tmp:
with httpx.stream("GET", url, timeout=300, follow_redirects=True) as resp:
with httpx.stream(
"GET", url, timeout=300, follow_redirects=True
) as resp:
resp.raise_for_status()
total = 0
for chunk in resp.iter_bytes(chunk_size=1024 * 1024):
Expand Down Expand Up @@ -317,7 +309,9 @@ def fetch_aws_pricing(regions: list[str], now: str) -> list[dict]:
continue

product_fields = parse_product(product)
product_desc = (attrs.get("usagetype", "") + " " + attrs.get("operation", "")).strip()
product_desc = (
attrs.get("usagetype", "") + " " + attrs.get("operation", "")
).strip()

for dim in _extract_ondemand_prices(terms, sku):
row = {
Expand Down Expand Up @@ -410,117 +404,39 @@ def fetch_claude_pricing(now: str) -> list[dict]:


# ---------------------------------------------------------------------------
# ROSA cluster instance discovery
# ROSA service fees — static rates from aws.amazon.com/rosa/pricing/
# ---------------------------------------------------------------------------


def _oc_get_contexts() -> list[dict]:
"""Get oc/kubectl contexts."""
try:
result = subprocess.run(
["oc", "config", "get-contexts", "-o", "name"],
capture_output=True,
text=True,
timeout=10,
)
if result.returncode != 0:
return []
names = [n.strip() for n in result.stdout.strip().splitlines() if n.strip()]
contexts = []
for name in names:
parts = name.split("/")
cluster_url = parts[1] if len(parts) > 1 else name
env = "unknown"
for env_name in ("prod", "stage", "staging", "uat", "dev"):
if env_name in cluster_url.lower():
env = env_name
break
contexts.append(
{
"name": name,
"cluster_url": cluster_url,
"environment": env,
}
)
return contexts
except (FileNotFoundError, subprocess.TimeoutExpired):
return []


def fetch_rosa_instances(now: str) -> list[dict]:
"""Discover instance types from ROSA clusters via oc."""
try:
result = subprocess.run(
["oc", "version", "--client"],
capture_output=True,
text=True,
timeout=10,
def fetch_rosa_service_fees(now: str) -> list[dict]:
"""Generate ROSA service fee rows from published AWS rates."""
rows: list[dict] = []
for fee in ROSA_SERVICE_FEES:
rows.append(
{
"provider": "aws",
"service": "rosa",
"region": "global",
"sku": fee["sku"],
"description": fee["description"],
"instance_type": None,
"instance_family": None,
"vcpu": None,
"memory_gb": None,
"storage_type": None,
"usage_type": fee["usage_type"],
"unit": fee["unit"],
"price_per_unit": fee["price_per_unit"],
"currency": "USD",
"effective_date": now[:10],
"model_name": None,
"tier_start": None,
"tier_end": None,
"fetched_at": now,
}
)
if result.returncode != 0:
print(" oc CLI not available, skipping ROSA discovery", file=sys.stderr)
return []
except (FileNotFoundError, subprocess.TimeoutExpired):
print(" oc CLI not available, skipping ROSA discovery", file=sys.stderr)
return []

contexts = _oc_get_contexts()
if not contexts:
print(" No oc contexts found, skipping ROSA discovery")
return []

all_rows: list[dict] = []
for ctx in contexts:
cluster_name = ctx["cluster_url"]
env = ctx["environment"]
print(f" Discovering nodes in {cluster_name} ({env})...")

try:
result = subprocess.run(
["oc", "get", "nodes", "--context", ctx["name"], "-o", "json"],
capture_output=True,
text=True,
timeout=30,
)
if result.returncode != 0:
print(f" ERROR: {result.stderr.strip()[:200]}", file=sys.stderr)
continue

nodes_data = json.loads(result.stdout)
for node in nodes_data.get("items", []):
labels = node.get("metadata", {}).get("labels", {})
node_name = node.get("metadata", {}).get("name", "")
instance_type = labels.get("node.kubernetes.io/instance-type", "unknown")
az = labels.get("topology.kubernetes.io/zone", "")

role = "worker"
for label_key in labels:
if "master" in label_key or "control-plane" in label_key:
role = "master"
break
if "infra" in label_key:
role = "infra"
break

all_rows.append(
{
"cluster_name": cluster_name,
"environment": env,
"node_name": node_name,
"instance_type": instance_type,
"role": role,
"availability_zone": az,
"fetched_at": now,
}
)

print(f" {len(nodes_data.get('items', []))} nodes")

except json.JSONDecodeError:
print(" ERROR: failed to parse node JSON", file=sys.stderr)
except subprocess.TimeoutExpired:
print(f" ERROR: oc timed out for {cluster_name}", file=sys.stderr)

return all_rows
print(f" {len(rows)} ROSA service fee entries")
return rows


# ---------------------------------------------------------------------------
Expand Down Expand Up @@ -564,41 +480,19 @@ def upsert_pricing(conn: sqlite3.Connection, rows: list[dict]) -> int:
return count


def upsert_rosa_instances(conn: sqlite3.Connection, rows: list[dict]) -> int:
"""Upsert ROSA cluster instance rows."""
count = 0
for row in rows:
conn.execute(
"""INSERT INTO rosa_cluster_instance (
cluster_name, environment, node_name, instance_type,
role, availability_zone, fetched_at
) VALUES (
:cluster_name, :environment, :node_name, :instance_type,
:role, :availability_zone, :fetched_at
)
ON CONFLICT(cluster_name, node_name) DO UPDATE SET
environment=excluded.environment,
instance_type=excluded.instance_type,
role=excluded.role,
availability_zone=excluded.availability_zone,
fetched_at=excluded.fetched_at""",
row,
)
count += 1
return count


# ---------------------------------------------------------------------------
# Main
# ---------------------------------------------------------------------------


def main() -> None:
parser = argparse.ArgumentParser(description="Fetch cloud pricing into pricing.db")
parser.add_argument("--aws-only", action="store_true", help="Fetch AWS pricing only")
parser.add_argument("--claude-only", action="store_true", help="Fetch Claude pricing only")
parser.add_argument("--rosa-only", action="store_true", help="ROSA discovery only")
parser.add_argument("--skip-rosa", action="store_true", help="Skip ROSA discovery")
parser.add_argument(
"--aws-only", action="store_true", help="Fetch AWS pricing only"
)
parser.add_argument(
"--claude-only", action="store_true", help="Fetch Claude pricing only"
)
parser.add_argument(
"--regions",
nargs="+",
Expand All @@ -614,9 +508,8 @@ def main() -> None:
args = parser.parse_args()

# Determine what to fetch
fetch_aws = not args.claude_only and not args.rosa_only
fetch_claude = not args.aws_only and not args.rosa_only
fetch_rosa = not args.aws_only and not args.claude_only and not args.skip_rosa
fetch_aws = not args.claude_only
fetch_claude = not args.aws_only

conn = init_db()
now = datetime.now(timezone.utc).isoformat()
Expand Down Expand Up @@ -663,21 +556,14 @@ def main() -> None:
(now,),
)

# --- ROSA ---
if fetch_rosa:
print("\n=== ROSA Cluster Discovery ===")
rows = fetch_rosa_instances(now)
# --- ROSA service fees ---
if fetch_aws or (not args.claude_only):
print("\n=== ROSA Service Fees (aws.amazon.com/rosa/pricing) ===")
rows = fetch_rosa_service_fees(now)
if rows:
count = upsert_rosa_instances(conn, rows)
count = upsert_pricing(conn, rows)
conn.commit()
print(f" Total: {count} cluster nodes upserted")
conn.execute(
"INSERT INTO _meta (key, value) VALUES ('last_rosa_fetch', ?) "
"ON CONFLICT(key) DO UPDATE SET value=excluded.value",
(now,),
)
else:
errors.append("ROSA: no nodes discovered (oc may not be available)")
print(f" Total: {count} ROSA fee entries upserted")
Comment on lines +559 to +566
Copy link
Copy Markdown

@coderabbitai coderabbitai Bot May 20, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🛠️ Refactor suggestion | 🟠 Major | ⚡ Quick win

Redundant condition — simplify to if fetch_aws:

Since fetch_aws = not args.claude_only (line 511), the condition fetch_aws or (not args.claude_only) is logically equivalent to just fetch_aws. The second disjunct is always equal to the first.

♻️ Proposed fix 
-        # --- ROSA service fees ---
-        if fetch_aws or (not args.claude_only):
+        # --- ROSA service fees (bundled with AWS) ---
+        if fetch_aws:
             print("\n=== ROSA Service Fees (aws.amazon.com/rosa/pricing) ===")
             rows = fetch_rosa_service_fees(now)
📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change
# --- ROSA service fees ---
if fetch_aws or (not args.claude_only):
print("\n=== ROSA Service Fees (aws.amazon.com/rosa/pricing) ===")
rows = fetch_rosa_service_fees(now)
if rows:
count = upsert_rosa_instances(conn, rows)
count = upsert_pricing(conn, rows)
conn.commit()
print(f" Total: {count} cluster nodes upserted")
conn.execute(
"INSERT INTO _meta (key, value) VALUES ('last_rosa_fetch', ?) "
"ON CONFLICT(key) DO UPDATE SET value=excluded.value",
(now,),
)
else:
errors.append("ROSA: no nodes discovered (oc may not be available)")
print(f" Total: {count} ROSA fee entries upserted")
# --- ROSA service fees (bundled with AWS) ---
if fetch_aws:
print("\n=== ROSA Service Fees (aws.amazon.com/rosa/pricing) ===")
rows = fetch_rosa_service_fees(now)
if rows:
count = upsert_pricing(conn, rows)
conn.commit()
print(f" Total: {count} ROSA fee entries upserted")
🤖 Prompt for AI Agents 
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@scripts/fetch_pricing.py` around lines 559 - 566, The if-condition for ROSA
service fees is redundant because fetch_aws is already set as not
args.claude_only; replace the compound condition "if fetch_aws or (not
args.claude_only)" with the simpler "if fetch_aws" where the ROSA block calls
fetch_rosa_service_fees(...) and upserts via upsert_pricing(...), leaving the
rest of the block unchanged.


conn.commit()

Expand Down
Loading
Loading