_gcp

sc_crawler.vendors._gcp

Functions:

Name	Description
inventory_compliance_frameworks	Manual list of compliance frameworks known for GCP.
inventory_regions	List all available GCP regions via API calls.
inventory_zones	List all available GCP zones via API calls.
inventory_servers	List all available GCP servers available in all zones.
inventory_server_prices	List all available GCP server ondemand prices in all regions.
inventory_server_prices_spot	List all available GCP server spot prices in all regions.
inventory_storages	List all available GCP disk storage options available in all zones.
inventory_storage_prices	List all available GCP disk storage prices in all regions.
inventory_traffic_prices	List inbound and outbound network traffic prices in all GCP regions.
inventory_ipv4_prices	List the price of an attached IPv4 address in all GCP regions.
inventory_databases	List all available GCP Cloud SQL for PostgreSQL database types via API calls.
inventory_database_prices	List all available GCP Cloud SQL for PostgreSQL on-demand prices in all regions.
inventory_database_storages	List all available GCP Cloud SQL for PostgreSQL storage options.
inventory_database_storage_prices	List all available GCP Cloud SQL for PostgreSQL storage prices in all regions.

inventory_compliance_frameworks

inventory_compliance_frameworks(vendor)

Manual list of compliance frameworks known for GCP.

Resources: https://cloud.google.com/compliance?hl=en

Source code in sc_crawler/vendors/_gcp.py

def inventory_compliance_frameworks(vendor):
    """Manual list of compliance frameworks known for GCP.

    Resources: <https://cloud.google.com/compliance?hl=en>"""
    return map_compliance_frameworks_to_vendor(
        vendor.vendor_id, ["hipaa", "soc2t2", "iso27001"]
    )

inventory_regions

inventory_regions(vendor)

List all available GCP regions via API calls.

Some data sources are not available from APIs, and were collected manually:

• location: https://cloud.google.com/compute/docs/regions-zones#available and https://en.wikipedia.org/wiki/Google_data_centers,
• lon/lat coordinates: https://en.wikipedia.org/wiki/Google_data_centers#Locations and approximation based on the city when no more accurate data was available.
• energy carbon data: https://cloud.google.com/sustainability/region-carbon#data and https://github.com/GoogleCloudPlatform/region-carbon-info,
• launch dates were collected from Wikipedia and GCP blog posts, such as https://medium.com/@retomeier/an-annotated-history-of-googles-cloud-platform-90b90f948920 and https://cloud.google.com/blog/products/infrastructure/introducing-new-google-cloud-regions.

Note that many GCP regions use more than 90% green energy, but the related flag in our database is set to False as not being 100%.

Source code in sc_crawler/vendors/_gcp.py

def inventory_regions(vendor):
    """List all available GCP regions via API calls.

    Some data sources are not available from APIs, and were collected manually:

    - location: <https://cloud.google.com/compute/docs/regions-zones#available> and <https://en.wikipedia.org/wiki/Google_data_centers>,
    - lon/lat coordinates: <https://en.wikipedia.org/wiki/Google_data_centers#Locations> and approximation based on the city when no more accurate data was available.
    - energy carbon data: <https://cloud.google.com/sustainability/region-carbon#data> and <https://github.com/GoogleCloudPlatform/region-carbon-info>,
    - launch dates were collected from [Wikipedia](https://en.wikipedia.org/wiki/Google_Cloud_Platform#Regions_and_zones) and GCP blog posts, such as <https://medium.com/@retomeier/an-annotated-history-of-googles-cloud-platform-90b90f948920> and <https://cloud.google.com/blog/products/infrastructure/introducing-new-google-cloud-regions>.

    Note that many GCP regions use more than 90% green energy,
    but the related flag in our database is set to `False` as not being 100%.
    """

    manual_data = {
        "africa-south1": {
            "country_id": "ZA",
            "city": "Johannesburg",
            # https://cloud.google.com/blog/products/infrastructure/heita-south-africa-new-cloud-region
            "founding_year": 2024,
            "green_energy": False,
            # approximation based on city
            "lat": -26.0420631,
            "lon": 28.0589808,
        },
        "asia-east1": {
            "country_id": "TW",
            "state": "Changhua County",
            "founding_year": 2013,
            "green_energy": False,
            "lat": 24.1385,
            "lon": 120.425722,
        },
        "asia-east2": {
            "country_id": "HK",
            # https://cloud.google.com/blog/products/gcp/gcps-region-in-hong-kong-is-now-open
            "founding_year": 2018,
            "green_energy": False,
            # approximation based on country
            "lat": 22.2772377,
            "lon": 114.1703066,
            "display_name": "Hong Kong",
        },
        "asia-northeast1": {
            "country_id": "JP",
            "city": "Tokyo",
            "state": "Japan",
            "founding_year": 2016,
            "green_energy": False,
            # approximation based on city
            "lat": 35.6433846,
            "lon": 139.7684933,
        },
        "asia-northeast2": {
            "country_id": "JP",
            "city": "Osaka",
            "founding_year": 2019,
            "green_energy": False,
            # approximation based on city
            "lat": 34.6696646,
            "lon": 135.4846612,
        },
        "asia-northeast3": {
            "country_id": "KR",
            "city": "Seoul",
            "founding_year": 2020,
            "green_energy": False,
            # approximation based on city
            "lat": 37.5514982,
            "lon": 126.97784,
        },
        "asia-south1": {
            "country_id": "IN",
            "city": "Mumbai",
            "founding_year": 2017,
            "green_energy": False,
            # approximation based on city
            "lat": 19.0709441,
            "lon": 72.8726468,
        },
        "asia-south2": {
            "country_id": "IN",
            "city": "Delhi",
            "founding_year": 2021,
            "green_energy": False,
            # approximation based on city
            "lat": 28.6439839,
            "lon": 76.9284239,
        },
        "asia-southeast1": {
            "country_id": "SG",
            "city": "Jurong West",
            "founding_year": 2017,
            "green_energy": False,
            "lat": 1.351333,
            "lon": 103.709778,
        },
        "asia-southeast2": {
            "country_id": "ID",
            "city": "Jakarta",
            "founding_year": 2020,
            "green_energy": False,
            # approximation based on city
            "lat": -6.2297401,
            "lon": 106.747117,
        },
        "asia-southeast3": {
            "country_id": "TH",
            "city": "Bangkok",
            "founding_year": 2025,
            "green_energy": False,
            # approximation based on city
            "lat": 15.870032,
            "lon": 100.992538,
        },
        "australia-southeast1": {
            "country_id": "AU",
            "city": "Sydney",
            "founding_year": 2017,
            "green_energy": False,
            # approximation based on city
            "lat": -33.8375583,
            "lon": 150.9488095,
        },
        "australia-southeast2": {
            "country_id": "AU",
            "city": "Melbourne",
            "founding_year": 2021,
            "green_energy": False,
            # approximation based on city
            "lat": -37.8038607,
            "lon": 144.7119569,
        },
        "europe-central2": {
            "country_id": "PL",
            "city": "Warsaw",
            "founding_year": 2021,
            "green_energy": False,
            # approximation based on city
            "lat": 52.2328871,
            "lon": 20.8966164,
        },
        "europe-north1": {
            "country_id": "FI",
            "city": "Hamina",
            "founding_year": 2018,
            "green_energy": False,
            "lat": 60.536578,
            "lon": 27.117003,
        },
        "europe-north2": {
            "country_id": "SE",
            "city": "Stockholm",
            "founding_year": 2025,
            "green_energy": False,
            # approximation based on city
            "lat": 59.334591,
            "lon": 18.06324,
        },
        "europe-southwest1": {
            "country_id": "ES",
            "city": "Madrid",
            "founding_year": 2022,
            "green_energy": False,
            "lat": 40.519533,
            "lon": -3.340937,
        },
        "europe-west1": {
            "country_id": "BE",
            "city": "St. Ghislain",
            # https://medium.com/@retomeier/an-annotated-history-of-googles-cloud-platform-90b90f948920
            "founding_year": 2015,
            "green_energy": False,
            "lat": 50.469333,
            "lon": 3.865472,
        },
        "europe-west10": {
            "country_id": "DE",
            "city": "Berlin",
            "founding_year": 2023,
            "green_energy": False,
            # approximation based on city
            "lat": 52.5105672,
            "lon": 13.3806972,
        },
        "europe-west12": {
            "country_id": "IT",
            "city": "Turin",
            "founding_year": 2023,
            "green_energy": False,
            "lat": 45.146729,
            "lon": 7.742147,
        },
        "europe-west2": {
            "country_id": "GB",
            "city": "London",
            "founding_year": 2017,
            "green_energy": False,
            # approximation based on city
            "lat": 51.5090133,
            "lon": -0.2118157,
        },
        "europe-west3": {
            "country_id": "DE",
            "city": "Frankfurt",
            "founding_year": 2017,
            "green_energy": False,
            "lat": 50.12263,
            "lon": 8.974168,
        },
        "europe-west4": {
            "country_id": "NL",
            "city": "Eemshaven",
            "founding_year": 2018,
            "green_energy": False,
            "lat": 52.790105,
            "lon": 5.029219,
        },
        "europe-west6": {
            "country_id": "CH",
            "city": "Zurich",
            "founding_year": 2019,
            "green_energy": False,
            "lat": 47.445926,
            "lon": 8.210909,
        },
        "europe-west8": {
            "country_id": "IT",
            "city": "Milan",
            "founding_year": 2022,
            "green_energy": False,
            # approximation based on city
            "lat": 45.4615551,
            "lon": 9.1389572,
        },
        "europe-west9": {
            "country_id": "FR",
            "city": "Paris",
            "founding_year": 2022,
            "green_energy": False,
            # approximation based on city
            "lat": 48.8641797,
            "lon": 2.3109137,
        },
        "me-central1": {
            "country_id": "QA",
            "city": "Doha",
            "founding_year": 2023,
            "green_energy": False,
            # approximation based on city
            "lat": 25.272868,
            "lon": 51.4717522,
        },
        "me-central2": {
            "country_id": "SA",
            "city": "Dammam",
            "founding_year": 2023,
            "green_energy": False,
            # approximation based on city
            "lat": 26.3826288,
            "lon": 49.9675732,
        },
        "me-west1": {
            "country_id": "IL",
            "city": "Tel Aviv",
            "founding_year": 2022,
            "green_energy": False,
            # approximation based on city
            "lat": 32.0491183,
            "lon": 34.7891105,
        },
        "northamerica-northeast1": {
            "country_id": "CA",
            "city": "Montréal",
            "founding_year": 2018,
            "green_energy": True,
            # approximation based on city
            "lat": 45.4933996,
            "lon": -73.728239,
        },
        "northamerica-northeast2": {
            "country_id": "CA",
            "city": "Toronto",
            "founding_year": 2021,
            "green_energy": False,
            # approximation based on city
            "lat": 43.72666,
            "lon": -79.5355309,
        },
        "southamerica-east1": {
            "country_id": "BR",
            "city": "Osasco",
            "state": "São Paulo",
            "founding_year": 2017,
            "green_energy": False,
            # approximation based on city
            "lat": -23.5267431,
            "lon": -46.8096539,
        },
        "southamerica-west1": {
            "country_id": "CL",
            "city": "Santiago",
            "founding_year": 2021,
            "green_energy": False,
            "lat": -33.520515,
            "lon": -70.721695,
        },
        # NOTE this is not announced yet, but showing up in API from time to time
        "northamerica-south1": {
            # https://mexicobusiness.news/cloudanddata/news/google-cloud-announces-first-mexican-data-region-queretaro
            "country_id": "MX",
            "city": "Queretaro",
            "founding_year": 2025,
            "green_energy": False,
            # approximation based on city
            "lat": 20.5896,
            "lon": -100.3897,
        },
        "us-central1": {
            "country_id": "US",
            "city": "Council Bluffs",
            "state": "Iowa",
            "founding_year": 2009,
            "green_energy": False,
            "lat": 41.168253,
            "lon": -95.796125,
        },
        "us-east1": {
            "country_id": "US",
            "city": "Moncks Corner",
            "state": "South Carolina",
            "founding_year": 2015,
            "green_energy": False,
            "lat": 33.064111,
            "lon": -80.043361,
        },
        "us-east4": {
            "country_id": "US",
            "city": "Ashburn",
            "state": "Virginia",
            "founding_year": 2017,
            "green_energy": False,
            "lat": 38.943331,
            "lon": -77.524336,
        },
        "us-east5": {
            "country_id": "US",
            "city": "Columbus",
            "state": "Ohio",
            "founding_year": 2022,
            "green_energy": False,
            # approximation based on city
            "lat": 39.9773124,
            "lon": -83.0423282,
        },
        "us-south1": {
            "country_id": "US",
            "city": "Dallas",
            "state": "Texas",
            "founding_year": 2022,
            "green_energy": False,
            "lat": 32.44317,
            "lon": -97.062324,
        },
        "us-west1": {
            "country_id": "US",
            "city": "The Dalles",
            "state": "Oregon",
            "founding_year": 2016,
            "green_energy": False,
            "lat": 45.632511,
            "lon": -121.202267,
        },
        "us-west2": {
            "country_id": "US",
            "city": "Los Angeles",
            "state": "California",
            "founding_year": 2018,
            "green_energy": False,
            # approximation based on city
            "lat": 34.0549694,
            "lon": -118.3753618,
        },
        "us-west3": {
            "country_id": "US",
            "city": "Salt Lake City",
            "state": "Utah",
            "founding_year": 2020,
            "green_energy": False,
            # approximation based on city
            "lat": 40.7386099,
            "lon": -111.9609998,
        },
        "us-west4": {
            "country_id": "US",
            "city": "Las Vegas",
            "state": "Nevada",
            "founding_year": 2020,
            "green_energy": False,
            "lat": 36.055625,
            "lon": -115.010226,
        },
    }

    # add API reference and display names
    for k, v in manual_data.items():
        v["api_reference"] = k
        if v.get("display_name") is None:
            v["display_name"] = v.get("city", v.get("state", ""))
            if v.get("display_name"):
                v["display_name"] = v["display_name"] + " (" + v["country_id"] + ")"
            else:
                v["display_name"] = v["country_id"]

    regions = _regions()
    items = []
    for region in regions:
        with sentry_capture_or_raise(vendor=vendor):
            if region.name not in manual_data:
                raise KeyError(f"Unknown region metadata for {region.name}")
            item = {
                "vendor_id": vendor.vendor_id,
                "region_id": str(region.id),
                "name": region.name,
            }
            for k, v in manual_data[region.name].items():
                item[k] = v
            items.append(item)
    return items

inventory_zones

inventory_zones(vendor)

List all available GCP zones via API calls.

Source code in sc_crawler/vendors/_gcp.py

def inventory_zones(vendor):
    """List all available GCP zones via API calls."""
    items = []
    regions = scmodels_to_dict(vendor.regions, keys=["name"])
    for zone in _zones():
        items.append(
            {
                "vendor_id": vendor.vendor_id,
                # example `zone.region`:
                # https://www.googleapis.com/compute/v1/projects/algebraic-pier-412621/regions/us-east4
                "region_id": regions[zone.region.split("/")[-1]].region_id,
                "zone_id": str(zone.id),
                "name": zone.name,
                "api_reference": zone.name,
                "display_name": zone.name,
            }
        )
    return items

inventory_servers

inventory_servers(vendor)

List all available GCP servers available in all zones.

Lifecycle (machineTypes.deprecated.state): DEPRECATED -> PLANNED_FOR_RETIREMENT, OBSOLETE / DELETED -> RETIRED, empty / ACTIVE -> ACTIVE. See _gcp_machine_type_status.

Source code in sc_crawler/vendors/_gcp.py

def inventory_servers(vendor):
    """List all available GCP servers available in all zones.

    Lifecycle (`machineTypes.deprecated.state`): DEPRECATED -> PLANNED_FOR_RETIREMENT,
    OBSOLETE / DELETED -> RETIRED, empty / ACTIVE -> ACTIVE.
    See `_gcp_machine_type_status`.
    """
    servers = parallel_fetch_servers(vendor, _search_servers, "name", "zones")
    servers = preprocess_servers(servers, vendor, add_vendor_id)
    return servers

inventory_server_prices

inventory_server_prices(vendor)

List all available GCP server ondemand prices in all regions.

Prices cover the predefined vCPU, memory, attached GPUs, and bundled Local SSD (from machineTypes.bundledLocalSsds).

Source code in sc_crawler/vendors/_gcp.py

def inventory_server_prices(vendor):
    """List all available GCP server ondemand prices in all regions.

    Prices cover the predefined vCPU, memory, attached GPUs, and bundled Local
    SSD (from `machineTypes.bundledLocalSsds`).
    """
    return _inventory_server_prices(vendor, Allocation.ONDEMAND)

inventory_server_prices_spot

inventory_server_prices_spot(vendor)

List all available GCP server spot prices in all regions.

Same components as inventory_server_prices (including bundled Local SSD).

Source code in sc_crawler/vendors/_gcp.py

def inventory_server_prices_spot(vendor):
    """List all available GCP server spot prices in all regions.

    Same components as `inventory_server_prices` (including bundled Local SSD).
    """
    return _inventory_server_prices(vendor, Allocation.SPOT)

inventory_storages

inventory_storages(vendor)

List all available GCP disk storage options available in all zones.

For more details on the disk types, check https://cloud.google.com/compute/docs/disks#disk-types.

Source code in sc_crawler/vendors/_gcp.py

def inventory_storages(vendor):
    """List all available GCP disk storage options available in all zones.

    For more details on the disk types, check <https://cloud.google.com/compute/docs/disks#disk-types>."""
    vendor.progress_tracker.start_task(
        name="Scanning zone(s) for storage(s)", total=len(vendor.zones)
    )

    def search_storages(zone: Zone, vendor: Vendor) -> List[dict]:
        zone_storages = []
        for storage in _storages(zone.name):
            valid_sizes = storage.valid_disk_size.replace("GB", "").split("-")
            zone_storages.append(
                {
                    "storage_id": str(storage.id),
                    "vendor_id": vendor.vendor_id,
                    "name": storage.name,
                    "description": storage.description,
                    "storage_type": (
                        StorageType.SSD
                        if storage.name != "pd-standard"
                        else StorageType.HDD
                    ),
                    "max_iops": None,
                    "max_throughput": None,
                    "min_size": int(valid_sizes[0]),
                    "max_size": int(valid_sizes[1]),
                }
            )
        vendor.log(f"{len(zone_storages)} storage(s) found in {zone.name}.")
        vendor.progress_tracker.advance_task()
        return zone_storages

    with ThreadPoolExecutor(max_workers=8) as executor:
        storages = executor.map(search_storages, vendor.zones, repeat(vendor))
    storages = list(chain.from_iterable(storages))

    vendor.log(f"{len(storages)} storage(s) found in {len(vendor.zones)} zones.")
    storages = list({p["name"]: p for p in storages}.values())
    vendor.log(f"{len(storages)} unique storage(s) found.")
    storages = [s for s in storages if s["name"] in STORAGE_ALLOWLIST]
    vendor.log(f"{len(storages)} storage(s) after dropping items with complex pricing.")
    vendor.progress_tracker.hide_task()
    return storages

inventory_storage_prices

inventory_storage_prices(vendor)

List all available GCP disk storage prices in all regions.

Source code in sc_crawler/vendors/_gcp.py

def inventory_storage_prices(vendor):
    """List all available GCP disk storage prices in all regions."""
    regions = scmodels_to_dict(vendor.regions, keys=["name"])
    skus = _skus_dict()
    items = []
    for storage in vendor.storages:
        storage_regions = skus["storage"][storage.name].keys()
        for storage_region in storage_regions:
            # skip edge regions
            region = regions.get(storage_region)
            if region is None:
                vendor.log(
                    f"Skip unknown '{storage_region}' region for {storage.name}",
                    DEBUG,
                )
                continue

            price, currency = skus["storage"][storage.name][storage_region]["ondemand"]
            for zone in region.zones:
                items.append(
                    {
                        "vendor_id": vendor.vendor_id,
                        "region_id": region.region_id,
                        "storage_id": storage.storage_id,
                        "unit": PriceUnit.GB_MONTH,
                        "price": float(price),
                        "currency": currency,
                    }
                )
    return items

inventory_traffic_prices

inventory_traffic_prices(vendor)

List inbound and outbound network traffic prices in all GCP regions.

Source code in sc_crawler/vendors/_gcp.py

def inventory_traffic_prices(vendor):
    """List inbound and outbound network traffic prices in all GCP regions."""
    regions = scmodels_to_dict(vendor.regions, keys=["name"])
    skus = _skus("Compute Engine")
    items = []
    for sku in skus:
        # skip not processed items early
        if sku.category.resource_family != "Network":
            continue
        if sku.category.resource_group not in [
            "StandardInternetEgress",
            "StandardInternetIngress",
        ]:
            continue

        # helper variables
        traffic_regions = sku.service_regions
        tiered_rates = sku.pricing_info[0].pricing_expression.tiered_rates
        price_tiers = []
        for i in range(len(tiered_rates)):
            price_tiers.append(
                {
                    "lower": tiered_rates[i].start_usage_amount,
                    "upper": "Infinity"
                    if i == len(tiered_rates) - 1
                    else tiered_rates[i + 1].start_usage_amount,
                    "price": tiered_rates[i].unit_price.nanos / 1e9,
                }
            )

        for traffic_region in traffic_regions:
            region = regions.get(traffic_region)
            if region is None:
                vendor.log(
                    f"Skip unknown '{traffic_region}' region for {sku.description}",
                    DEBUG,
                )
                continue
            items.append(
                {
                    "vendor_id": vendor.vendor_id,
                    "region_id": region.region_id,
                    "price": max([float(t["price"]) for t in price_tiers]),
                    "price_tiered": price_tiers,
                    "currency": tiered_rates[0].unit_price.currency_code,
                    "unit": PriceUnit.GB_MONTH,
                    "direction": (
                        TrafficDirection.OUT
                        if sku.category.resource_group == "StandardInternetEgress"
                        else TrafficDirection.IN
                    ),
                }
            )

    return items

inventory_ipv4_prices

inventory_ipv4_prices(vendor)

List the price of an attached IPv4 address in all GCP regions.

Note that this data was not found using the APIs (only unattached static IPs), so the values are recorded manually from https://cloud.google.com/vpc/network-pricing#ipaddress.

Source code in sc_crawler/vendors/_gcp.py

def inventory_ipv4_prices(vendor):
    """List the price of an attached IPv4 address in all GCP regions.

    Note that this data was not found using the APIs (only unattached static IPs),
    so the values are recorded manually from <https://cloud.google.com/vpc/network-pricing#ipaddress>.
    """
    # skus = _skus("Compute Engine")
    # for sku in skus:
    #     if sku.category.resource_family != "Network":
    #         continue
    #     if sku.description == "Static Ip Charge":
    #         pass
    items = []
    for region in vendor.regions:
        items.append(
            {
                "vendor_id": vendor.vendor_id,
                "region_id": region.region_id,
                "price": 0.005,
                "currency": "USD",
                "unit": PriceUnit.HOUR,
            }
        )
    return items

inventory_databases

inventory_databases(vendor)

List all available GCP Cloud SQL for PostgreSQL database types via API calls.

Source code in sc_crawler/vendors/_gcp.py

def inventory_databases(vendor):
    """List all available GCP Cloud SQL for PostgreSQL database types via API calls."""
    vendor.progress_tracker.start_task(name="Fetching PostgreSQL tier(s)", total=None)
    meta = _pg_sqladmin_metadata()
    _, ha_families = _pg_billing_catalog()
    vendor.progress_tracker.hide_task()

    rows = []
    tiers = meta["tiers"]
    vendor.progress_tracker.start_task(name="Processing database(s)", total=len(tiers))
    for tier in tiers:
        tier_name = tier.get("tier")
        if not tier_name:
            vendor.progress_tracker.advance_task()
            continue
        # Skip custom shapes (db-custom-*), catalog predefined tiers only.
        # https://cloud.google.com/sql/docs/postgres/machine-series-overview
        if tier_name.startswith("db-custom-"):
            vendor.progress_tracker.advance_task()
            continue

        if match := _PG_NAMED_TIER_CPU_RE.search(tier_name):
            cpu_count = int(match.group(1))
        else:
            cpu_count = None

        ram_bytes = int(tier.get("RAM") or 0)
        memory_amount = int(ram_bytes / 1_048_576) if ram_bytes else None

        stripped = tier_name.removeprefix("db-")
        parts = stripped.split("-")
        family_slug = (
            "-".join(parts[:-1]) if parts and parts[-1].isdigit() else stripped
        )

        spec_parts: list[str] = []
        if cpu_count is not None:
            spec_parts.append(f"{cpu_count} vCPU{'s' if cpu_count != 1 else ''}")
        if memory_amount is not None:
            memory_gib = round(memory_amount / 1024, 1)
            gib_label = (
                f"{int(memory_gib)} GB RAM"
                if memory_gib == int(memory_gib)
                else f"{memory_gib:g} GB RAM"
            )
            spec_parts.append(gib_label)
        family_label = _PG_TIER_FAMILY_LABELS.get(
            family_slug, family_slug.replace("-", " ").title()
        )
        description = f"PostgreSQL Cloud SQL {family_label}"
        if spec_parts:
            description = f"{description} ({', '.join(spec_parts)})"
        server = next(
            (
                item
                for item in vendor.servers
                if item.api_reference == tier_name.removeprefix("db-")
            ),
            None,
        )
        server_id = server.server_id if server is not None else None
        status = server.status if server is not None else Status.ACTIVE
        raw_regions = tier.get("region")
        if isinstance(raw_regions, str):
            tier_regions = [raw_regions]
        elif isinstance(raw_regions, list):
            tier_regions = [region for region in raw_regions if isinstance(region, str)]
        else:
            tier_regions = []

        sku_family = _pg_sku_family(tier_name)
        is_enterprise_plus = sku_family.startswith("enterprise_plus")
        engine_versions = [
            version
            for version in meta["engine_versions"]
            if version in _PG_SUPPORTED_MAJOR_VERSIONS
        ]
        if is_enterprise_plus:
            plus_versions = _PG_ENTERPRISE_PLUS_MAJOR_VERSIONS
            if sku_family == "enterprise_plus_c4a":
                # PostgreSQL 12 is unsupported on C4A only, not on N2 / C4.
                # https://cloud.google.com/sql/docs/postgres/machine-series-overview
                plus_versions = plus_versions - {"12"}
            engine_versions = [
                version for version in engine_versions if version in plus_versions
            ]

        ha: list[DatabaseHaLevel] = []
        ha_strategy: list[DatabaseHaStrategy] = []
        has_regional_ha = False
        if tier_regions:
            # https://cloud.google.com/sql/docs/postgres/high-availability
            # Regional billing meters -> same-region multi-zone HA (standby).
            # Zonal billing meters -> non-HA.
            # Enterprise Plus Advanced DR -> cross-region replica promotion (not multi-region HA).
            has_regional_ha = any(
                (region, sku_family) in ha_families for region in tier_regions
            )
            if has_regional_ha:
                ha.append(DatabaseHaLevel.MULTI_ZONE)
                ha_strategy.append(DatabaseHaStrategy.PASSIVE_STANDBY)
        ha.append(DatabaseHaLevel.NONE)
        ha_strategy.append(DatabaseHaStrategy.NONE)

        disk_quota_bytes = int(tier.get("DiskQuota") or 0)
        storage_extra_max = (
            disk_quota_bytes // 1_000_000_000 if disk_quota_bytes else None
        )

        rows.append(
            {
                "vendor_id": vendor.vendor_id,
                "database_id": tier_name,
                "name": tier_name,
                "api_reference": tier_name,
                # https://www.pulumi.com/registry/packages/gcp/api-docs/sql/databaseinstance/
                # Machine size lives under settings (Pulumi DatabaseInstance.settings.tier).
                "api_reference_object": {"settings": {"tier": tier_name}},
                "display_name": tier_name,
                "description": description,
                # TODO: not clear which are the server instances for for perf-optimized and memory-optimized
                # db instances (like db-perf-optimized-N-2, db-memory-optimized-N-4, etc.)
                "server_id": server_id,
                "engine": DatabaseEngine.POSTGRESQL,
                "wire_protocol": DatabaseWireProtocol.POSTGRESQL,
                "engine_versions": engine_versions,
                "family": family_slug,
                "vcpus": cpu_count,
                "memory_amount": memory_amount,
                "storage_size": None,
                "ha": DatabaseHaLevel.ordered(ha),
                "ha_strategy": DatabaseHaStrategy.ordered(ha_strategy),
                # https://cloud.google.com/sql/docs/postgres/configure-ip
                # https://cloud.google.com/sql/docs/postgres/private-ip
                # https://cloud.google.com/sql/docs/postgres/configure-private-service-connect
                # https://cloud.google.com/sql/docs/postgres/iam-authentication
                # https://cloud.google.com/sql/docs/postgres/configure-ssl-instance
                # https://cloud.google.com/sql/docs/postgres/cmek
                # https://cloud.google.com/sql/docs/postgres/pg-audit
                # Authorized networks, private IP (PSA tenant VPC), PSC, IAM DB auth,
                # sslMode (ENCRYPTED_ONLY / client certs), CMEK, pgaudit.
                "security_features": [
                    DatabaseSecurityFeature.IP_FILTERING,
                    DatabaseSecurityFeature.PRIVATE_NETWORK,
                    DatabaseSecurityFeature.NETWORK_PEERING,
                    DatabaseSecurityFeature.IDENTITY_BASED_AUTH,
                    DatabaseSecurityFeature.CLIENT_CERT_AUTH,
                    DatabaseSecurityFeature.ENFORCED_TLS,
                    DatabaseSecurityFeature.CUSTOMER_MANAGED_KEYS,
                    DatabaseSecurityFeature.AUDIT_LOGGING,
                ],
                # https://cloud.google.com/sql/docs/postgres/admin-api/rest/v1/instances
                # Minimum data disk size is 10 GB.
                "storage_extra_min": 10,
                "storage_extra_max": storage_extra_max,
                # https://cloud.google.com/sql/docs/postgres/instance-settings
                "storage_extra_autosize": True,
                # https://cloud.google.com/sql/docs/postgres/cmek
                "disk_encryption": True,
                # https://cloud.google.com/sql/docs/postgres/maintenance
                # Minor versions are applied via Cloud SQL maintenance updates.
                "auto_upgrade_versions": True,
                # https://cloud.google.com/sql/docs/postgres/backup-recovery/backups
                "scheduled_backups": True,
                # https://cloud.google.com/sql/docs/postgres/backup-recovery/configure-pitr
                # Max transactionLogRetentionDays for PITR by edition; not in tiers API.
                # Enterprise: 1-7 days; Enterprise Plus: 1-35 days.
                "continuous_backups": (35 if is_enterprise_plus else 7),
                "custom_config": meta["custom_config"],
                "custom_extensions": meta["custom_extensions"],
                # https://cloud.google.com/sql/docs/postgres/replication
                # Shared-core tiers do not support read replicas.
                # Up to 10 replicas per primary (cascading also allowed).
                "max_read_replicas": (0 if tier_name in _PG_SHARED_TIERS else 10),
                # https://cloud.google.com/sql/docs/postgres/managed-connection-pooling
                "connection_pool": True,
                # https://cloud.google.com/sql/docs/postgres/monitor-instance
                "system_monitoring": True,
                # https://cloud.google.com/sql/docs/postgres/using-query-insights
                "database_monitoring": True,
                # https://cloud.google.com/sql/docs/postgres/use-index-advisor
                # Index advisor recommends CREATE INDEX (Enterprise Plus only); operator applies.
                "autotuning_advice": is_enterprise_plus,
                "autotuning_apply": False,
                # https://cloud.google.com/sql/sla
                # Enterprise Plus + HA: 99.99%; Enterprise + HA: 99.95%.
                # Shared-core and zonal (non-HA) instances are excluded from the SLA.
                "sla": (
                    99.99
                    if is_enterprise_plus and has_regional_ha
                    else (
                        99.95
                        if DatabaseHaLevel.MULTI_ZONE in ha
                        and tier_name not in _PG_SHARED_TIERS
                        else None
                    )
                ),
                "status": status,
            }
        )
        vendor.progress_tracker.advance_task()
    vendor.progress_tracker.hide_task()
    return rows

inventory_database_prices

inventory_database_prices(vendor)

List all available GCP Cloud SQL for PostgreSQL on-demand prices in all regions.

Source code in sc_crawler/vendors/_gcp.py

def inventory_database_prices(vendor):
    """List all available GCP Cloud SQL for PostgreSQL on-demand prices in all regions."""
    vendor.progress_tracker.start_task(name="Fetching Cloud SQL SKU(s)", total=None)
    compute_index, _ = _pg_billing_catalog()
    tiers = _pg_sqladmin_metadata()["tiers"]
    vendor.progress_tracker.hide_task()

    items = []
    vendor.progress_tracker.start_task(
        name="Processing database_price(s)", total=len(tiers)
    )
    for tier in tiers:
        tier_name = tier.get("tier")
        if not tier_name:
            vendor.progress_tracker.advance_task()
            continue
        if tier_name.startswith("db-custom-"):
            vendor.progress_tracker.advance_task()
            continue

        if match := _PG_NAMED_TIER_CPU_RE.search(tier_name):
            cpu_count = int(match.group(1))
        else:
            cpu_count = None

        ram_bytes = int(tier.get("RAM") or 0)
        memory_gib = ram_bytes / (1024**3) if ram_bytes else None

        raw_regions = tier.get("region")
        if isinstance(raw_regions, str):
            tier_regions = {raw_regions}
        elif isinstance(raw_regions, list):
            tier_regions = {region for region in raw_regions if isinstance(region, str)}
        else:
            tier_regions = set()

        sku_family = _pg_sku_family(tier_name)
        availabilities = ("zonal", "regional")

        for region in vendor.regions:
            if tier_regions and region.api_reference not in tier_regions:
                continue

            for availability in availabilities:
                hourly = currency = None
                if sku_family == "shared":
                    component = _PG_SHARED_TIERS[tier_name]
                    instance_sku = compute_index.get(
                        (region.api_reference, "shared", component, availability)
                    )
                    if instance_sku is not None:
                        hourly = _sku_unit_price(instance_sku)
                        if hourly is not None:
                            tiered = instance_sku.pricing_info[
                                0
                            ].pricing_expression.tiered_rates
                            currency = tiered[0].unit_price.currency_code or "USD"
                elif cpu_count is not None and memory_gib is not None:
                    vcpu_sku = compute_index.get(
                        (region.api_reference, sku_family, "vcpu", availability)
                    )
                    ram_sku = compute_index.get(
                        (region.api_reference, sku_family, "ram", availability)
                    )
                    if vcpu_sku is not None and ram_sku is not None:
                        vcpu_hourly = _sku_unit_price(vcpu_sku)
                        ram_hourly = _sku_unit_price(ram_sku)
                        if vcpu_hourly is not None and ram_hourly is not None:
                            hourly = vcpu_hourly * cpu_count + ram_hourly * memory_gib
                            vcpu_tiered = vcpu_sku.pricing_info[
                                0
                            ].pricing_expression.tiered_rates
                            currency = vcpu_tiered[0].unit_price.currency_code or "USD"

                if hourly is None:
                    continue
                # https://cloud.google.com/sql/docs/postgres/high-availability
                # Zonal = standalone; Regional HA = multi-zone standby (exactly 2x on
                # the pricing page; we use Regional billing SKUs, not a multiplier).
                if availability == "regional":
                    ha = DatabaseHaLevel.MULTI_ZONE
                    ha_strategy = DatabaseHaStrategy.PASSIVE_STANDBY
                else:
                    ha = DatabaseHaLevel.NONE
                    ha_strategy = DatabaseHaStrategy.NONE
                items.append(
                    {
                        "vendor_id": vendor.vendor_id,
                        "region_id": region.region_id,
                        "database_id": tier_name,
                        "allocation": Allocation.ONDEMAND,
                        "ha": ha,
                        "ha_strategy": ha_strategy,
                        "unit": PriceUnit.HOUR,
                        "price": hourly,
                        "price_upfront": 0,
                        "price_tiered": [],
                        "currency": currency or "USD",
                    }
                )
        vendor.progress_tracker.advance_task()
    vendor.progress_tracker.hide_task()
    return items

inventory_database_storages

inventory_database_storages(vendor)

List all available GCP Cloud SQL for PostgreSQL storage options.

Source code in sc_crawler/vendors/_gcp.py

def inventory_database_storages(vendor):
    """List all available GCP Cloud SQL for PostgreSQL storage options."""
    vendor.progress_tracker.start_task(name="Fetching Cloud SQL SKU(s)", total=None)
    found = {
        storage_id
        for sku in _cloud_sql_skus()
        if (storage_id := _pg_storage_id(sku.description or ""))
    }
    vendor.progress_tracker.hide_task()

    items = []
    vendor.progress_tracker.start_task(
        name="Processing database_storage(s)", total=len(found)
    )
    for storage_id in sorted(found):
        spec = _PG_STORAGE_SPECS[storage_id]
        items.append(
            {
                "vendor_id": vendor.vendor_id,
                "database_storage_id": storage_id,
                "name": spec["name"],
                "description": spec["description"],
                "scope": DatabaseStorageScope.DATA,
                "min_size": spec["min_size"],
                "max_size": spec["max_size"],
                "max_iops": spec["max_iops"],
                "max_throughput": spec["max_throughput"],
            }
        )
        vendor.progress_tracker.advance_task()
    vendor.progress_tracker.hide_task()
    return items

inventory_database_storage_prices

inventory_database_storage_prices(vendor)

List all available GCP Cloud SQL for PostgreSQL storage prices in all regions.

Source code in sc_crawler/vendors/_gcp.py

def inventory_database_storage_prices(vendor):
    """List all available GCP Cloud SQL for PostgreSQL storage prices in all regions."""
    vendor.progress_tracker.start_task(name="Fetching Cloud SQL SKU(s)", total=None)
    skus = _cloud_sql_skus()
    vendor.progress_tracker.hide_task()

    items = []
    seen: set[tuple[str, str]] = set()
    vendor.progress_tracker.start_task(
        name="Processing database_storage_price(s)", total=len(skus)
    )
    for sku in skus:
        storage_id = _pg_storage_id(sku.description or "")
        if not storage_id:
            vendor.progress_tracker.advance_task()
            continue

        unit_price = _sku_unit_price(sku)
        if unit_price is None:
            vendor.progress_tracker.advance_task()
            continue
        usage_unit = (
            sku.pricing_info[0].pricing_expression.usage_unit
            if sku.pricing_info
            else ""
        )
        if usage_unit == "GiBy.mo":
            monthly = unit_price
        elif usage_unit == "GiBy.h":
            monthly = unit_price * 730
        else:
            vendor.progress_tracker.advance_task()
            continue

        tiered = sku.pricing_info[0].pricing_expression.tiered_rates
        currency = tiered[0].unit_price.currency_code or "USD"
        for region in vendor.regions:
            if region.api_reference not in sku.service_regions:
                continue
            key = (region.region_id, storage_id)
            if key in seen:
                continue
            seen.add(key)
            items.append(
                {
                    "vendor_id": vendor.vendor_id,
                    "region_id": region.region_id,
                    "database_storage_id": storage_id,
                    "unit": PriceUnit.GB_MONTH,
                    "price": monthly,
                    "price_upfront": 0,
                    "price_tiered": [],
                    "currency": currency,
                }
            )
        vendor.progress_tracker.advance_task()
    vendor.progress_tracker.hide_task()
    return items