From 01ef8d6b38b62d25c0940020c345b542f65ee81c Mon Sep 17 00:00:00 2001
From: Ben Zhang <ben.zhang@watonomous.ca>
Date: Mon, 14 Oct 2024 03:15:53 +0000
Subject: [PATCH] Implement periodic GC

---
 README.md               |  2 +-
 server/requirements.txt |  3 ++-
 server/src/main.py      | 28 ++++++++++++++++++++--------
 3 files changed, 23 insertions(+), 10 deletions(-)

diff --git a/README.md b/README.md
index 3f197fa..39e2462 100644
--- a/README.md
+++ b/README.md
@@ -10,7 +10,7 @@ Coming soon:
 - [x] File upload API (we may be able to simply use the [publisher](https://cvmfs.readthedocs.io/en/stable/cpt-repository-gateway.html#publisher-configuration). It has nice features like being able to handle concurrent transactions.)
     - The publisher appears to be bottlenecked at 20MiB/s when running the server in Kubernetes, and around 80MiB/s when running in Docker. `iperf` gives much higher bandwidth (between nodes and between the Kubernetes container and nodes), so it's likely not a network bottleneck.
     - When using the custom FastAPI upload server, speeds reach over 400MiB/s easily. We'll adopt this approach.
-- [ ] Garbage collection
+- [x] Garbage collection
 - [ ] Better documentation
 - [ ] Automatic [whitelist re-signing](https://cvmfs.readthedocs.io/en/stable/apx-security.html#signature-details)
 
diff --git a/server/requirements.txt b/server/requirements.txt
index afc0a41..8b03d3a 100644
--- a/server/requirements.txt
+++ b/server/requirements.txt
@@ -1,4 +1,5 @@
 watcloud-utils @ git+https://github.com/WATonomous/watcloud-utils.git@c8ce1006716e65971f750560f90f442721b3777d
 python-slugify>=8.0.4,<9
 python-multipart>=0.0.12,<1
-uvicorn>=0.31.1,<1
\ No newline at end of file
+uvicorn>=0.31.1,<1
+apscheduler>=3.10.4,<4
\ No newline at end of file
diff --git a/server/src/main.py b/server/src/main.py
index fe06b01..0fe3a23 100644
--- a/server/src/main.py
+++ b/server/src/main.py
@@ -4,16 +4,19 @@
 import subprocess
 import sys
 import time
+from contextlib import asynccontextmanager
 from pathlib import Path
 from threading import Lock
 
 import typer
 import uvicorn
+from apscheduler.schedulers.background import BackgroundScheduler
+from apscheduler.triggers.cron import CronTrigger
 from fastapi import HTTPException, UploadFile
 from fastapi.responses import FileResponse
 from slugify import slugify
 from typing_extensions import Annotated
-from watcloud_utils.fastapi import WATcloudFastAPI
+from watcloud_utils.fastapi import WATcloudFastAPI, FastAPI
 from watcloud_utils.logging import logger, set_up_logging
 from watcloud_utils.typer import app
 
@@ -99,13 +102,22 @@ def init_cvmfs_repo(
     print(f"Successfully initialized CVMFS repo: {repo_name}")
     print(f"The public key is available via HTTP at GET /cvmfs-meta/{repo_name}.pub")
 
-@app.command()
-def start_server():
-    print("Starting server")
-    while True:
-        pass
-
-fastapi_app = WATcloudFastAPI(logger=logger)
+@asynccontextmanager
+async def fastapi_lifespan(app: FastAPI):
+    """
+    This function wraps the FastAPI app in a lifespan context manager.
+    i.e. it allows us to run code when the app starts and stops.
+    """
+    try:
+        scheduler.start()
+        # Run garbage collection every minute
+        scheduler.add_job(gc, CronTrigger.from_crontab("* * * * *"))
+        yield
+    finally:
+        scheduler.shutdown()
+
+scheduler = BackgroundScheduler()
+fastapi_app = WATcloudFastAPI(logger=logger, lifespan=fastapi_lifespan)
 transaction_lock = Lock()
 
 @fastapi_app.post("/repos/{repo_name}")