Files
Oxicloud/src/infrastructure/services/retry_blob_backend.rs
T
Edouard Vanbelle 465fbe2480 feat(errors): classify transient failures on the type, not by string-matching
Step 1 of docs/plan/jobs-handling-recoverable-error.md, and the blocker
for the rest of it: the engine cannot retry-then-pause until it can tell
"the provider is down" from "this data is wrong". Both arrived as
`ErrorKind::InternalError`, so the distinction survived only inside a
formatted message.

`RetryBlobBackend` was reading that message. Literally:

    let msg = err.to_string().to_lowercase();
    msg.contains("timeout") || msg.contains("503") || msg.contains("reset by peer")

Fragile in a specific way — an SDK reformatting its `Display` turns
retrying off with nothing failing to say so — and blind to any status
code that never made it into the text.

Adds `ErrorKind::TransientBackend` and `DomainError::is_transient()`.
One predicate, so the retry decorator and the job engine cannot classify
the same failure differently. `Timeout` counts (transient by
construction); everything else must say so explicitly. The default is
"not retryable" because that fails visibly, whereas retrying a permanent
fault burns attempts and — once the engine wires this up — holds
`migration_readonly` while it does.

A kind rather than a `transient: bool` field: 21 struct-literal sites
construct `DomainError` directly and would all have needed touching for
a change that is conceptually about classification. The plan allowed
either.

`s3_domain_error` does the classification where the status is still in
hand. Transient: 5xx, 429, and the SlowDown / RequestTimeout /
ThrottlingException codes that arrive as 400 (status alone is not
enough), plus dispatch-level I/O and timeouts. Permanent: other 4xx —
credentials, missing bucket, malformed request — and construction
failures. `ResponseError` counts as transient since truncation on the
wire is the usual cause and the attempt cap bounds being wrong.

Applied at the five S3 sites that wrap an SDK error, including
`ListObjectsV2` — `backend_consistency` fails the whole run on an
enumeration error, so a throttle midway through a large bucket should be
retryable rather than discarding the sweep.

The exhaustive `ErrorKind` match in `interfaces/errors.rs` forced the
HTTP decision, which is the right friction: 503, not 500. The request
was fine and may succeed shortly, which is what a caller needs to decide
whether to retry and what a proxy keys off to avoid caching the failure.

The substring matcher stays for now, behind the typed check, with the
deletion condition written down: it goes when every backend wrapping a
remote SDK error classifies at the point of wrapping. Removing it before
then would silently reduce retrying on the unconverted backends, which
is the worse direction. Azure is the one left, and it is queued for the
official-SDK migration anyway.

Not yet wired: `RunOutcome::PausedRetryable` (step 2) and the engine's
bounded backoff (step 3). This commit only makes the distinction
representable.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-08 06:23:24 +02:00

410 lines
13 KiB
Rust
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
//! `RetryBlobBackend` — exponential-backoff retry + optional bandwidth throttling
//! decorator for remote blob backends.
//!
//! Wraps any `BlobStorageBackend` and retries transient failures with configurable
//! exponential backoff. Optionally throttles upload/download bandwidth via
//! inter-chunk sleeps.
use std::path::{Path, PathBuf};
use std::pin::Pin;
use std::sync::Arc;
use std::time::Duration;
use crate::application::ports::blob_storage_ports::{
BlobStorageBackend, BlobStream, StorageHealthStatus,
};
use crate::domain::errors::DomainError;
use bytes::Bytes;
// ── Retry policy ───────────────────────────────────────────────────
/// Exponential backoff retry configuration.
#[derive(Debug, Clone)]
pub struct RetryPolicy {
/// Maximum number of retry attempts (0 = no retries).
pub max_retries: u32,
/// Initial backoff duration before the first retry.
pub initial_backoff: Duration,
/// Maximum backoff duration (capped).
pub max_backoff: Duration,
/// Multiplier applied to backoff after each attempt.
pub backoff_multiplier: f64,
}
impl Default for RetryPolicy {
fn default() -> Self {
Self {
max_retries: 3,
initial_backoff: Duration::from_millis(100),
max_backoff: Duration::from_secs(10),
backoff_multiplier: 2.0,
}
}
}
// ── RetryBlobBackend ───────────────────────────────────────────────
/// Decorator that retries failed backend operations with exponential backoff.
pub struct RetryBlobBackend {
inner: Arc<dyn BlobStorageBackend>,
policy: RetryPolicy,
}
impl RetryBlobBackend {
pub fn new(inner: Arc<dyn BlobStorageBackend>, policy: RetryPolicy) -> Self {
Self { inner, policy }
}
}
/// Execute an async closure with exponential backoff retry.
///
/// `name` is a lazy label: the success path (the overwhelmingly common
/// case) never materializes it, so per-op `format!("op({hash})")`
/// allocations only happen on an actual retry (benches/ROUND11.md §14:
/// 64.5 → 0.7 ns, −2 allocs per blob op).
async fn retry_async<F, Fut, T, L>(
policy: &RetryPolicy,
name: L,
mut f: F,
) -> Result<T, DomainError>
where
F: FnMut() -> Fut,
Fut: std::future::Future<Output = Result<T, DomainError>>,
L: Fn() -> String,
{
let mut attempt = 0u32;
let mut backoff = policy.initial_backoff;
loop {
match f().await {
Ok(v) => return Ok(v),
Err(e) if attempt < policy.max_retries && is_retryable(&e) => {
attempt += 1;
tracing::warn!(
"Retry {}/{} for {} after error: {} (backoff {:?})",
attempt,
policy.max_retries,
name(),
e,
backoff
);
tokio::time::sleep(backoff).await;
let next =
Duration::from_secs_f64(backoff.as_secs_f64() * policy.backoff_multiplier);
backoff = next.min(policy.max_backoff);
}
Err(e) => return Err(e),
}
}
}
/// Determine if an error is likely transient (network timeout, 5xx, etc.).
///
/// Asks the error first. `DomainError::is_transient` is the single
/// answer to that question, so this decorator and the job engine cannot
/// classify the same failure differently.
///
/// **The substring arm is transitional.** It is what this function used
/// to be, in full: a `to_lowercase()` scan of `Display` output for
/// "timeout", "503", "reset by peer" and friends. That is fragile in a
/// specific way — an SDK reformatting its error text silently turns
/// retries off, with nothing failing to say so — and it cannot see a
/// status code that never made it into the message.
///
/// It stays only until every backend classifies at the point of
/// wrapping, where the status is still in hand. Deleting it before then
/// would silently REDUCE retrying on the backends not yet converted,
/// which is the worse direction to be wrong in. Delete it once
/// `grep -rn "transient_backend" src/infrastructure/services/` covers
/// every backend that wraps a remote SDK error.
fn is_retryable(err: &DomainError) -> bool {
if err.is_transient() {
return true;
}
let msg = err.to_string().to_lowercase();
msg.contains("timeout")
|| msg.contains("connection")
|| msg.contains("503")
|| msg.contains("500")
|| msg.contains("429")
|| msg.contains("temporarily")
|| msg.contains("broken pipe")
|| msg.contains("reset by peer")
}
impl BlobStorageBackend for RetryBlobBackend {
fn initialize(
&self,
) -> Pin<Box<dyn std::future::Future<Output = Result<(), DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
Box::pin(async move {
retry_async(
&policy,
|| "initialize".to_string(),
|| {
let inner = inner.clone();
async move { inner.initialize().await }
},
)
.await
})
}
fn put_blob(
&self,
hash: &str,
source_path: &Path,
) -> Pin<Box<dyn std::future::Future<Output = Result<u64, DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
let path = source_path.to_path_buf();
Box::pin(async move {
retry_async(
&policy,
|| format!("put_blob({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
let path = path.clone();
async move { inner.put_blob(&hash, &path).await }
},
)
.await
})
}
fn put_blob_from_bytes(
&self,
hash: &str,
data: Bytes,
) -> Pin<Box<dyn std::future::Future<Output = Result<u64, DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("put_blob_from_bytes({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
let data = data.clone();
async move { inner.put_blob_from_bytes(&hash, data).await }
},
)
.await
})
}
// Without this override the trait default would re-route the CDC chunk
// write through `put_blob_from_bytes` above — reinstating the remote
// backend's exists-probe (HEAD/get_properties) per chunk that the
// `_unsynced` fast path exists to skip.
fn put_blob_from_bytes_unsynced(
&self,
hash: &str,
data: Bytes,
) -> Pin<Box<dyn std::future::Future<Output = Result<u64, DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("put_blob_from_bytes_unsynced({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
let data = data.clone();
async move { inner.put_blob_from_bytes_unsynced(&hash, data).await }
},
)
.await
})
}
// Forwarded WITHOUT retry wrapping: a failed fsync must surface, not be
// re-issued — after an fsync error the kernel may have dropped the dirty
// pages, so a retried fsync can report success for data that was lost.
fn sync_blobs(
&self,
hashes: &[String],
) -> Pin<Box<dyn std::future::Future<Output = Result<(), DomainError>> + Send + '_>> {
self.inner.sync_blobs(hashes)
}
fn get_blob_stream(
&self,
hash: &str,
) -> Pin<Box<dyn std::future::Future<Output = Result<BlobStream, DomainError>> + Send + '_>>
{
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("get_blob_stream({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
async move { inner.get_blob_stream(&hash).await }
},
)
.await
})
}
fn get_blob_range_stream(
&self,
hash: &str,
start: u64,
end: Option<u64>,
) -> Pin<Box<dyn std::future::Future<Output = Result<BlobStream, DomainError>> + Send + '_>>
{
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("get_blob_range({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
async move { inner.get_blob_range_stream(&hash, start, end).await }
},
)
.await
})
}
fn delete_blob(
&self,
hash: &str,
) -> Pin<Box<dyn std::future::Future<Output = Result<(), DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("delete_blob({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
async move { inner.delete_blob(&hash).await }
},
)
.await
})
}
fn blob_exists(
&self,
hash: &str,
) -> Pin<Box<dyn std::future::Future<Output = Result<bool, DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("blob_exists({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
async move { inner.blob_exists(&hash).await }
},
)
.await
})
}
fn blob_size(
&self,
hash: &str,
) -> Pin<Box<dyn std::future::Future<Output = Result<u64, DomainError>> + Send + '_>> {
let inner = self.inner.clone();
let policy = self.policy.clone();
let hash = hash.to_string();
Box::pin(async move {
retry_async(
&policy,
|| format!("blob_size({hash})"),
|| {
let inner = inner.clone();
let hash = hash.clone();
async move { inner.blob_size(&hash).await }
},
)
.await
})
}
fn health_check(
&self,
) -> Pin<
Box<dyn std::future::Future<Output = Result<StorageHealthStatus, DomainError>> + Send + '_>,
> {
let inner = self.inner.clone();
let policy = self.policy.clone();
Box::pin(async move {
retry_async(
&policy,
|| "health_check".to_string(),
|| {
let inner = inner.clone();
async move { inner.health_check().await }
},
)
.await
})
}
fn backend_type(&self) -> &'static str {
"retry"
}
/// Transparent wrapper: the inner backend serves the bytes.
fn read_prefetch(&self) -> usize {
self.inner.read_prefetch()
}
fn local_blob_path(&self, hash: &str) -> Option<PathBuf> {
self.inner.local_blob_path(hash)
}
/// Pass-through wrapper — forward, so an unwrap started above still
/// reaches the cache. Inheriting the `None` default would silently
/// end the search at this layer and leave verification reading
/// through the cache after all.
fn uncached(&self) -> Option<Arc<dyn BlobStorageBackend>> {
self.inner.uncached()
}
/// Enumeration delegates to inner. Retry semantics apply per
/// call, not per batch — a single list call that fails after
/// exhausting retries surfaces the error to the tenant, which
/// treats it as a transient backend issue and skips the batch.
fn list_blob_hashes(
&self,
cursor: Option<String>,
limit: usize,
) -> Pin<
Box<
dyn std::future::Future<
Output = Result<
crate::application::ports::blob_storage_ports::BlobListPage,
DomainError,
>,
> + Send
+ '_,
>,
> {
self.inner.list_blob_hashes(cursor, limit)
}
}