Cloud Run のエフェメラルディスク機能を使ってみる #GoogleCloudNext

# ダミーファイル作成
dd if=/dev/zero of=testfile.bin bs=1M count=1024

# バケット作成
gcloud storage buckets create gs://${GOOGLE_CLOUD_PROJECT}-ephemeral-test --location=asia-northeast1

# ファイルをアップロード
gcloud storage cp testfile.bin gs://${GOOGLE_CLOUD_PROJECT}-ephemeral-test/testfile.bin

# サービスアカウントを作成
gcloud iam service-accounts create cr-ephemeral-test-sa \
  --display-name="Cloud Run Ephemeral Disk Test SA"

# バケットへの読み取り権限を付与
gcloud storage buckets add-iam-policy-binding gs://${GOOGLE_CLOUD_PROJECT}-ephemeral-test \
  --member="serviceAccount:cr-ephemeral-test-sa@${GOOGLE_CLOUD_PROJECT}.iam.gserviceaccount.com" \
  --role="roles/storage.objectViewer"

import os
import subprocess
import time
from flask import Flask, jsonify
from google.cloud import storage

app = Flask(__name__)

BUCKET_NAME = os.environ["BUCKET_NAME"]
FILE_NAME = "testfile.bin"
DOWNLOAD_PATH = os.environ.get("DOWNLOAD_PATH", f"/tmp/{FILE_NAME}")
BENCH_SIZE_MB = 512


def get_df(path):
    result = subprocess.run(["df", "-m", path], capture_output=True, text=True)
    lines = result.stdout.strip().split("\n")
    fields = lines[1].split()
    return {
        "filesystem": fields[0],
        "total_mb": int(fields[1]),
        "used_mb": int(fields[2]),
        "available_mb": int(fields[3]),
    }


def benchmark_io(directory):
    bench_file = os.path.join(directory, "bench.bin")
    chunk = b"\x00" * (1024 * 1024)

    start = time.time()
    with open(bench_file, "wb") as f:
        for _ in range(BENCH_SIZE_MB):
            f.write(chunk)
        f.flush()
        os.fsync(f.fileno())
    write_mbps = round(BENCH_SIZE_MB / (time.time() - start), 1)

    start = time.time()
    with open(bench_file, "rb") as f:
        while f.read(1024 * 1024):
            pass
    read_mbps = round(BENCH_SIZE_MB / (time.time() - start), 1)

    os.remove(bench_file)
    return {"write_mbps": write_mbps, "read_mbps": read_mbps}


@app.route("/download")
def download():
    client = storage.Client()
    bucket = client.bucket(BUCKET_NAME)
    blob = bucket.blob(FILE_NAME)
    blob.download_to_filename(DOWNLOAD_PATH)

    download_dir = os.path.dirname(DOWNLOAD_PATH)
    return jsonify({
        "download_path": DOWNLOAD_PATH,
        "filesystem": get_df(download_dir),
    })


@app.route("/benchmark")
def benchmark():
    download_dir = os.path.dirname(DOWNLOAD_PATH)
    return jsonify({
        "directory": download_dir,
        "io_benchmark": benchmark_io(download_dir),
    })


@app.route("/")
def index():
    return "ok"


if __name__ == "__main__":
    app.run(host="0.0.0.0", port=8080)

flask
google-cloud-storage

FROM python:3.12-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY app.py .
CMD ["python", "app.py"]

gcloud run deploy cr-tmp-test \
  --source . \
  --region asia-northeast1 \
  --execution-environment gen2 \
  --memory 2Gi \
  --service-account="cr-ephemeral-test-sa@${GOOGLE_CLOUD_PROJECT}.iam.gserviceaccount.com" \
  --set-env-vars "BUCKET_NAME=${GOOGLE_CLOUD_PROJECT}-ephemeral-test,DOWNLOAD_PATH=/tmp/testfile.bin" \
  --allow-unauthenticated

gcloud beta run deploy cr-disk-test \
  --source . \
  --region asia-northeast1 \
  --execution-environment gen2 \
  --memory 512Mi \
  --min-instances=0 \
  --max-instances=3 \
  --service-account="cr-ephemeral-test-sa@${GOOGLE_CLOUD_PROJECT}.iam.gserviceaccount.com" \
  --add-volume name=data,type=ephemeral-disk,size=10Gi \
  --add-volume-mount volume=data,mount-path=/mnt/data \
  --set-env-vars "BUCKET_NAME=${GOOGLE_CLOUD_PROJECT}-ephemeral-test,DOWNLOAD_PATH=/mnt/data/testfile.bin" \
  --allow-unauthenticated

# /tmp パターン
TMP_URL=$(gcloud run services describe cr-tmp-test --region asia-northeast1 --format='value(status.url)')
curl -H "Authorization: Bearer $(gcloud auth print-identity-token)" ${TMP_URL}/download

# エフェメラルディスクパターン
DISK_URL=$(gcloud run services describe cr-disk-test --region asia-northeast1 --format='value(status.url)')
curl -H "Authorization: Bearer $(gcloud auth print-identity-token)" ${DISK_URL}/download

curl -H "Authorization: Bearer $(gcloud auth print-identity-token)" ${DISK_URL}/benchmark

{"directory":"/mnt/data","io_benchmark":{"read_mbps":1749.5,"write_mbps":978.9}}