feat(tvix/store/pathinfo): add a Cache combinator r/8144

This allows querying two PathInfoService implementations sequentially,
and inserts into the "near" one if it's not there yet.

There is no negative cache, and put / listing is not implemented (for
now).

Change-Id: I24c3d0e0c3c2f0524a6cc7b2f3cbc33eb20cf92b
Reviewed-on: https://cl.tvl.fyi/c/depot/+/11636
Autosubmit: flokli <flokli@flokli.de>
Tested-by: BuildkiteCI
Reviewed-by: Connor Brewster <cbrewster@hey.com>

2 files changed, 113 insertions, 0 deletions

diff --git a/tvix/store/src/pathinfoservice/combinators.rs b/tvix/store/src/pathinfoservice/combinators.rs
new file mode 100644
index 000000000000..664144ef494b
--- /dev/null
+++ b/tvix/store/src/pathinfoservice/combinators.rs
@@ -0,0 +1,111 @@
+use crate::proto::PathInfo;
+use futures::stream::BoxStream;
+use nix_compat::nixbase32;
+use tonic::async_trait;
+use tracing::{debug, instrument};
+use tvix_castore::Error;
+
+use super::PathInfoService;
+
+/// Asks near first, if not found, asks far.
+/// If found in there, returns it, and *inserts* it into
+/// near.
+/// There is no negative cache.
+/// Inserts and listings are not implemented for now.
+pub struct Cache<PS1, PS2> {
+    near: PS1,
+    far: PS2,
+}
+
+impl<PS1, PS2> Cache<PS1, PS2> {
+    pub fn new(near: PS1, far: PS2) -> Self {
+        Self { near, far }
+    }
+}
+
+#[async_trait]
+impl<PS1, PS2> PathInfoService for Cache<PS1, PS2>
+where
+    PS1: PathInfoService,
+    PS2: PathInfoService,
+{
+    #[instrument(level = "trace", skip_all, fields(path_info.digest = nixbase32::encode(&digest)))]
+    async fn get(&self, digest: [u8; 20]) -> Result<Option<PathInfo>, Error> {
+        match self.near.get(digest).await? {
+            Some(path_info) => {
+                debug!("serving from cache");
+                Ok(Some(path_info))
+            }
+            None => {
+                debug!("not found in near, asking remote…");
+                match self.far.get(digest).await? {
+                    None => Ok(None),
+                    Some(path_info) => {
+                        debug!("found in remote, adding to cache");
+                        self.near.put(path_info.clone()).await?;
+                        Ok(Some(path_info))
+                    }
+                }
+            }
+        }
+    }
+
+    async fn put(&self, _path_info: PathInfo) -> Result<PathInfo, Error> {
+        Err(Error::StorageError("unimplemented".to_string()))
+    }
+
+    fn list(&self) -> BoxStream<'static, Result<PathInfo, Error>> {
+        Box::pin(tokio_stream::once(Err(Error::StorageError(
+            "unimplemented".to_string(),
+        ))))
+    }
+}
+
+#[cfg(test)]
+mod test {
+    use std::num::NonZeroUsize;
+
+    use crate::{
+        pathinfoservice::{LruPathInfoService, MemoryPathInfoService, PathInfoService},
+        tests::fixtures::PATH_INFO_WITH_NARINFO,
+    };
+
+    const PATH_INFO_DIGEST: [u8; 20] = [0; 20];
+
+    /// Helper function setting up an instance of a "far" and "near"
+    /// PathInfoService.
+    async fn create_pathinfoservice() -> super::Cache<LruPathInfoService, MemoryPathInfoService> {
+        // Create an instance of a "far" PathInfoService.
+        let far = MemoryPathInfoService::default();
+
+        // … and an instance of a "near" PathInfoService.
+        let near = LruPathInfoService::with_capacity(NonZeroUsize::new(1).unwrap());
+
+        // create a Pathinfoservice combining the two and return it.
+        super::Cache::new(near, far)
+    }
+
+    /// Getting from the far backend is gonna insert it into the near one.
+    #[tokio::test]
+    async fn test_populate_cache() {
+        let svc = create_pathinfoservice().await;
+
+        // query the PathInfo, things should not be there.
+        assert!(svc.get(PATH_INFO_DIGEST).await.unwrap().is_none());
+
+        // insert it into the far one.
+        svc.far.put(PATH_INFO_WITH_NARINFO.clone()).await.unwrap();
+
+        // now try getting it again, it should succeed.
+        assert_eq!(
+            Some(PATH_INFO_WITH_NARINFO.clone()),
+            svc.get(PATH_INFO_DIGEST).await.unwrap()
+        );
+
+        // peek near, it should now be there.
+        assert_eq!(
+            Some(PATH_INFO_WITH_NARINFO.clone()),
+            svc.near.get(PATH_INFO_DIGEST).await.unwrap()
+        );
+    }
+}
diff --git a/tvix/store/src/pathinfoservice/mod.rs b/tvix/store/src/pathinfoservice/mod.rs
index 5f5da2bff296..574bcc0b8b88 100644
--- a/tvix/store/src/pathinfoservice/mod.rs
+++ b/tvix/store/src/pathinfoservice/mod.rs
@@ -1,3 +1,4 @@
+mod combinators;
 mod from_addr;
 mod grpc;
 mod lru;
@@ -17,6 +18,7 @@ use tvix_castore::Error;
 
 use crate::proto::PathInfo;
 
+pub use self::combinators::Cache as CachePathInfoService;
 pub use self::from_addr::from_addr;
 pub use self::grpc::GRPCPathInfoService;
 pub use self::lru::LruPathInfoService;