From cf08a3f7315143a98dd2f6aa75e09c1035554f22 Mon Sep 17 00:00:00 2001
From: Fintan Halpenny <fintan.halpenny@gmail.com>
Date: Tue, 31 Oct 2023 15:19:34 +0000
Subject: [PATCH] node: run git gc on fetch

It is desirable to keep the Git repositories in Radicle efficient by
ensuring objects are packed and unreachable objects are removed.

Add a process call to `git gc` to achieve this. Two flags are used for
garbage collection: `--auto` and `--prune`. Their details are outlined below:

       --auto
           With this option, git gc checks whether any housekeeping is
           required; if not, it exits without performing any work.

       --prune=<date>
           Prune loose objects older than date (default is 2 weeks
           ago, overridable by the config variable
           gc.pruneExpire). --prune=now prunes loose objects
           regardless of their age and increases the risk of
           corruption if another process is writing to the repository
           concurrently; see "NOTES" below. --prune is on by default.

For `--prune` the default used is `1.hours.ago`. This was chosen over
`now` since that may be too aggressive after just fetching new data.

Signed-off-by: Fintan Halpenny <fintan.halpenny@gmail.com>
X-Clacks-Overhead: GNU Terry Pratchett
---
 radicle-node/src/runtime.rs        |  1 +
 radicle-node/src/worker.rs         | 14 +++++++-
 radicle-node/src/worker/garbage.rs | 53 ++++++++++++++++++++++++++++++
 3 files changed, 67 insertions(+), 1 deletion(-)
 create mode 100644 radicle-node/src/worker/garbage.rs

diff --git a/radicle-node/src/runtime.rs b/radicle-node/src/runtime.rs
index d7a4b056..05c30d21 100644
--- a/radicle-node/src/runtime.rs
+++ b/radicle-node/src/runtime.rs
@@ -256,6 +256,7 @@ impl Runtime {
             limit: FetchLimit::default(),
             info: git::UserInfo { alias, key: nid },
             local: nid,
+            expiry: worker::garbage::Expiry::default(),
         };
         let pool = worker::Pool::with(
             worker_recv,
diff --git a/radicle-node/src/worker.rs b/radicle-node/src/worker.rs
index 36528269..148400c3 100644
--- a/radicle-node/src/worker.rs
+++ b/radicle-node/src/worker.rs
@@ -3,6 +3,7 @@ mod channels;
 mod upload_pack;
 
 pub mod fetch;
+pub mod garbage;
 
 use std::path::PathBuf;
 use std::{io, time};
@@ -162,6 +163,9 @@ pub struct FetchConfig {
     pub info: git::UserInfo,
     /// Public key of the local peer.
     pub local: crypto::PublicKey,
+    /// Configuration for `git gc` garbage collection. Defaults to `1
+    /// hour ago`.
+    pub expiry: garbage::Expiry,
 }
 
 /// A worker that replicates git objects.
@@ -279,6 +283,7 @@ impl Worker {
             limit,
             info,
             local,
+            expiry,
         } = &self.fetch_config;
         let tracking =
             tracking::Config::new(*policy, *scope, tracking::Store::reader(tracking_db)?);
@@ -296,8 +301,15 @@ impl Worker {
             blocked,
             channels,
         )?;
+        let result = handle.fetch(rid, &self.storage, *limit, remote, refs_at)?;
 
-        Ok(handle.fetch(rid, &self.storage, *limit, remote, refs_at)?)
+        if let Err(e) = garbage::collect(&self.storage, rid, *expiry) {
+            // N.b. ensure that `git gc` works in debug mode.
+            debug_assert!(false, "`git gc` failed: {e}");
+
+            log::warn!(target: "worker", "Failed to run `git gc`: {e}");
+        }
+        Ok(result)
     }
 }
 
diff --git a/radicle-node/src/worker/garbage.rs b/radicle-node/src/worker/garbage.rs
new file mode 100644
index 00000000..ec19b14f
--- /dev/null
+++ b/radicle-node/src/worker/garbage.rs
@@ -0,0 +1,53 @@
+use std::process::{Command, ExitStatus, Stdio};
+use std::{fmt, io};
+
+use radicle::prelude::Id;
+use radicle::storage::ReadStorage;
+
+/// Default expiry time for objects.
+pub const EXPIRY_DEFAULT: Expiry = Expiry::Hours(1);
+
+/// Expiry of objects for garbage collector.
+#[derive(Copy, Clone, Debug, PartialEq, Eq)]
+pub enum Expiry {
+    Now,
+    Seconds(usize),
+    Hours(usize),
+    Days(usize),
+    Weeks(usize),
+}
+
+impl Default for Expiry {
+    fn default() -> Self {
+        EXPIRY_DEFAULT
+    }
+}
+
+impl fmt::Display for Expiry {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            Self::Now => f.write_str("now"),
+            Self::Seconds(s) => write!(f, "{s}.seconds.ago"),
+            Self::Hours(s) => write!(f, "{s}.hours.ago"),
+            Self::Days(s) => write!(f, "{s}.days.ago"),
+            Self::Weeks(s) => write!(f, "{s}.weeks.ago"),
+        }
+    }
+}
+
+/// Run Git garbage collector.
+pub fn collect(storage: &impl ReadStorage, rid: Id, expiry: Expiry) -> io::Result<ExitStatus> {
+    let git_dir = storage.path_of(&rid);
+    let mut gc = Command::new("git");
+    gc.current_dir(git_dir)
+        .env_clear()
+        .envs(std::env::vars().filter(|(key, _)| key == "PATH" || key.starts_with("GIT_TRACE")))
+        .args(["gc", &format!("--prune={expiry}"), "--auto"])
+        .stdout(Stdio::piped())
+        .stdin(Stdio::piped())
+        .stderr(Stdio::inherit());
+    let mut child = gc.spawn()?;
+    let status = child.wait()?;
+
+    Ok(status)
+}