feat: write chunks directly to temp upload files

Completed uploads used to copy every staged chunk into a second file before
renaming the result into data/complete. That doubled write volume and required
peak disk space for both the chunk set and the final file.

Write each chunk directly into one private temp upload file at its final offset
instead. After a chunk write succeeds, record a tiny durable completion marker
for progress and resume scans. Completion now verifies the temp file length and
all markers, then renames the temp file into the completed upload directory.

Add UPL_TEMP_DIR and --temp-dir so operators can choose where upload metadata,
markers, and temp files live. The default remains data/staging, and docs call
out that the temp directory must be on the same filesystem as data/complete for
atomic promotion. The nginx example now aliases only the completed upload
directory, and the smoke test verifies that final-file alias.

This keeps the existing length-based validation model; it does not add per-chunk
hashing.

Test Plan:
- just check
- just nginx-smoke
- cargo clippy && cargo clippy --benches && cargo clippy --tests
- cargo +nightly fmt --all
- cargo clippy && cargo clippy --benches && cargo clippy --tests

Refs: none
This commit is contained in:
2026-05-30 18:10:05 +02:00
parent 428af52e2f
commit c072b93726
10 changed files with 232 additions and 101 deletions
+11 -9
View File
@@ -42,13 +42,16 @@ async fn stores_chunks_and_reports_progress() -> Result<(), Box<dyn std::error::
let progress = get_progress(&app, &upload.upload_id).await?;
assert_eq!(progress.completed_chunks, vec![0, 1]);
let chunk_path = temp_dir
.path()
.join("staging")
.join(&upload.upload_id)
.join("chunks")
.join("000000.part");
assert_eq!(tokio::fs::metadata(chunk_path).await?.len(), CHUNK_SIZE);
let upload_dir = temp_dir.path().join("staging").join(&upload.upload_id);
assert_eq!(
tokio::fs::metadata(upload_dir.join(".upload.tmp"))
.await?
.len(),
CHUNK_SIZE + 3
);
assert!(upload_dir.join("completed").join("000000.done").is_file());
assert!(upload_dir.join("completed").join("000001.done").is_file());
assert!(!upload_dir.join("chunks").exists());
Ok(())
}
@@ -84,8 +87,7 @@ async fn rejects_out_of_range_chunk_index() -> Result<(), Box<dyn std::error::Er
}
#[tokio::test]
async fn accepts_duplicate_chunk_when_existing_length_matches()
-> Result<(), Box<dyn std::error::Error>> {
async fn accepts_duplicate_completed_chunk() -> Result<(), Box<dyn std::error::Error>> {
let temp_dir = TempDir::new()?;
let app = test_app(temp_dir.path());
let upload = create_upload(&app, temp_dir.path(), 4).await?;
+4 -8
View File
@@ -98,18 +98,14 @@ async fn rejects_incomplete_upload() -> Result<(), Box<dyn std::error::Error>> {
}
#[tokio::test]
async fn rejects_corrupt_chunk_file() -> Result<(), Box<dyn std::error::Error>> {
async fn rejects_tampered_temp_upload_file() -> Result<(), Box<dyn std::error::Error>> {
let temp_dir = TempDir::new()?;
let app = test_app(temp_dir.path());
let upload = create_upload(&app, "corrupt.bin", 4).await?;
let chunk_path = temp_dir
.path()
.join("staging")
.join(&upload.upload_id)
.join("chunks")
.join("000000.part");
tokio::fs::write(chunk_path, b"bad").await?;
let upload_dir = temp_dir.path().join("staging").join(&upload.upload_id);
tokio::fs::write(upload_dir.join(".upload.tmp"), b"bad").await?;
tokio::fs::write(upload_dir.join("completed").join("000000.done"), b"").await?;
let response = app
.oneshot(empty_request(
+2 -1
View File
@@ -43,7 +43,8 @@ async fn creates_upload_metadata_on_disk() -> Result<(), Box<dyn std::error::Err
let upload_dir = temp_dir.path().join("staging").join(&response.upload_id);
let meta_path = upload_dir.join("meta.json");
assert!(upload_dir.join("chunks").is_dir());
assert!(upload_dir.join(".upload.tmp").is_file());
assert!(upload_dir.join("completed").is_dir());
assert!(temp_dir.path().join("complete").is_dir());
let meta: UploadMeta = serde_json::from_slice(&tokio::fs::read(meta_path).await?)?;