feat(tvix/store): add blobservice r/5894

This adds a BlobService trait, and an implementation for it using sled,
and one using a HashMap.

Change-Id: Id6bc1b629195d0b26fc503bd7d2dc9e43c41c317
Reviewed-on: https://cl.tvl.fyi/c/depot/+/8087
Tested-by: BuildkiteCI
Reviewed-by: tazjin <tazjin@tvl.su>

3 files changed, 145 insertions, 0 deletions

diff --git a/tvix/store/src/blobservice/memory.rs b/tvix/store/src/blobservice/memory.rs
new file mode 100644
index 000000000000..73028f8f3c63
--- /dev/null
+++ b/tvix/store/src/blobservice/memory.rs
@@ -0,0 +1,59 @@
+use data_encoding::BASE64;
+use std::{
+    collections::HashMap,
+    sync::{Arc, RwLock},
+};
+use tracing::instrument;
+
+use crate::{proto, Error};
+
+use super::BlobService;
+
+#[derive(Clone)]
+pub struct MemoryBlobService {
+    db: Arc<RwLock<HashMap<Vec<u8>, proto::BlobMeta>>>,
+}
+
+impl MemoryBlobService {
+    pub fn new() -> Self {
+        let db = Arc::new(RwLock::new(HashMap::default()));
+
+        Self { db }
+    }
+}
+
+impl BlobService for MemoryBlobService {
+    #[instrument(skip(self, req), fields(blob.digest=BASE64.encode(&req.digest)))]
+    fn stat(&self, req: &proto::StatBlobRequest) -> Result<Option<proto::BlobMeta>, Error> {
+        if req.include_bao {
+            todo!("not implemented yet")
+        }
+
+        let db = self.db.read().unwrap();
+        // if include_chunks is also false, the user only wants to know if the
+        // blob is present at all.
+        if !req.include_chunks {
+            Ok(if db.contains_key(&req.digest) {
+                Some(proto::BlobMeta::default())
+            } else {
+                None
+            })
+        } else {
+            match db.get(&req.digest) {
+                None => Ok(None),
+                Some(blob_meta) => Ok(Some(blob_meta.clone())),
+            }
+        }
+    }
+
+    #[instrument(skip(self, blob_meta, blob_digest), fields(blob.digest = BASE64.encode(blob_digest)))]
+    fn put(&self, blob_digest: &[u8], blob_meta: proto::BlobMeta) -> Result<(), Error> {
+        let mut db = self.db.write().unwrap();
+
+        db.insert(blob_digest.to_vec(), blob_meta);
+
+        Ok(())
+        // TODO: make sure all callers make sure the chunks exist.
+        // TODO: where should we calculate the bao?
+    }
+}
diff --git a/tvix/store/src/blobservice/mod.rs b/tvix/store/src/blobservice/mod.rs
new file mode 100644
index 000000000000..53e941795e7e
--- /dev/null
+++ b/tvix/store/src/blobservice/mod.rs
@@ -0,0 +1,20 @@
+use crate::{proto, Error};
+
+mod memory;
+mod sled;
+
+pub use self::memory::MemoryBlobService;
+pub use self::sled::SledBlobService;
+
+/// The base trait all BlobService services need to implement.
+/// It provides information about how a blob is chunked,
+/// and allows creating new blobs by creating a BlobMeta (referring to chunks
+/// in a [crate::chunkservice::ChunkService]).
+pub trait BlobService {
+    /// Retrieve chunking information for a given blob
+    fn stat(&self, req: &proto::StatBlobRequest) -> Result<Option<proto::BlobMeta>, Error>;
+
+    /// Insert chunking information for a given blob.
+    /// Implementations SHOULD make sure chunks referred do exist.
+    fn put(&self, blob_digest: &[u8], blob_meta: proto::BlobMeta) -> Result<(), Error>;
+}
diff --git a/tvix/store/src/blobservice/sled.rs b/tvix/store/src/blobservice/sled.rs
new file mode 100644
index 000000000000..729a520ac2d6
--- /dev/null
+++ b/tvix/store/src/blobservice/sled.rs
@@ -0,0 +1,66 @@
+use std::path::PathBuf;
+
+use data_encoding::BASE64;
+use prost::Message;
+use tracing::instrument;
+
+use crate::{proto, Error};
+
+use super::BlobService;
+
+#[derive(Clone)]
+pub struct SledBlobService {
+    db: sled::Db,
+}
+
+impl SledBlobService {
+    pub fn new(p: PathBuf) -> Result<Self, sled::Error> {
+        let config = sled::Config::default().use_compression(true).path(p);
+        let db = config.open()?;
+
+        Ok(Self { db })
+    }
+}
+
+impl BlobService for SledBlobService {
+    #[instrument(name = "SledBlobService::stat", skip(self, req), fields(blob.digest=BASE64.encode(&req.digest)))]
+    fn stat(&self, req: &proto::StatBlobRequest) -> Result<Option<proto::BlobMeta>, Error> {
+        if req.include_bao {
+            todo!("not implemented yet")
+        }
+
+        // if include_chunks is also false, the user only wants to know if the
+        // blob is present at all.
+        if !req.include_chunks {
+            match self.db.contains_key(&req.digest) {
+                Ok(false) => Ok(None),
+                Ok(true) => Ok(Some(proto::BlobMeta::default())),
+                Err(e) => Err(Error::StorageError(e.to_string())),
+            }
+        } else {
+            match self.db.get(&req.digest) {
+                Ok(None) => Ok(None),
+                Ok(Some(data)) => match proto::BlobMeta::decode(&*data) {
+                    Ok(blob_meta) => Ok(Some(blob_meta)),
+                    Err(e) => Err(Error::StorageError(format!(
+                        "unable to parse blobmeta message for blob {}: {}",
+                        BASE64.encode(&req.digest),
+                        e
+                    ))),
+                },
+                Err(e) => Err(Error::StorageError(e.to_string())),
+            }
+        }
+    }
+
+    #[instrument(name = "SledBlobService::put", skip(self, blob_meta, blob_digest), fields(blob.digest = BASE64.encode(blob_digest)))]
+    fn put(&self, blob_digest: &[u8], blob_meta: proto::BlobMeta) -> Result<(), Error> {
+        let result = self.db.insert(blob_digest, blob_meta.encode_to_vec());
+        if let Err(e) = result {
+            return Err(Error::StorageError(e.to_string()));
+        }
+        Ok(())
+        // TODO: make sure all callers make sure the chunks exist.
+        // TODO: where should we calculate the bao?
+    }
+}