From: Christian Ebner <c.ebner@proxmox.com>
To: pbs-devel@lists.proxmox.com
Subject: [pbs-devel] [PATCH proxmox-backup v3 16/23] datastore: add locking to protect against races on chunk insert for s3
Date: Wed, 5 Nov 2025 13:22:26 +0100 [thread overview]
Message-ID: <20251105122233.439382-17-c.ebner@proxmox.com> (raw)
In-Reply-To: <20251105122233.439382-1-c.ebner@proxmox.com>
Acquire the per-chunk file lock to get exclusive access to the chunk
on insert and make it possible to guarantee that s3 backend
operations and the local caches reflect the consistent state without
other operations such as verify chunk renaming or garbage collection
interfering.
Signed-off-by: Christian Ebner <c.ebner@proxmox.com>
---
changes since version 2:
- no changes
pbs-datastore/src/datastore.rs | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/pbs-datastore/src/datastore.rs b/pbs-datastore/src/datastore.rs
index 698486ff4..f869320d8 100644
--- a/pbs-datastore/src/datastore.rs
+++ b/pbs-datastore/src/datastore.rs
@@ -1919,6 +1919,10 @@ impl DataStore {
match backend {
DatastoreBackend::Filesystem => self.inner.chunk_store.insert_chunk(chunk, digest),
DatastoreBackend::S3(s3_client) => {
+ let _chunk_guard = self
+ .inner
+ .chunk_store
+ .lock_chunk(digest, CHUNK_LOCK_TIMEOUT)?;
let chunk_data: Bytes = chunk.raw_data().to_vec().into();
let chunk_size = chunk_data.len() as u64;
let object_key = crate::s3::object_key_from_digest(digest)?;
@@ -1939,6 +1943,10 @@ impl DataStore {
) -> Result<(bool, u64), Error> {
let chunk_data = chunk.raw_data();
let chunk_size = chunk_data.len() as u64;
+ let _chunk_guard = self
+ .inner
+ .chunk_store
+ .lock_chunk(digest, CHUNK_LOCK_TIMEOUT)?;
// Avoid re-upload to S3 if the chunk is either present in the in-memory LRU cache
// or the chunk marker file exists on filesystem. The latter means the chunk has
--
2.47.3
_______________________________________________
pbs-devel mailing list
pbs-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pbs-devel
next prev parent reply other threads:[~2025-11-05 12:22 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-11-05 12:22 [pbs-devel] [PATCH proxmox-backup v3 00/23] fix chunk upload/insert, rename corrupt chunks and GC race conditions for s3 backend Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 01/23] sync: pull: instantiate backend only once per sync job Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 02/23] api/datastore: move group notes setting to the datastore Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 03/23] api/datastore: move snapshot deletion into dedicated datastore helper Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 04/23] api/datastore: move backup log upload by implementing " Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 05/23] api: backup: use datastore add_blob helper for backup session Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 06/23] api/datastore: add dedicated datastore helper to set snapshot notes Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 07/23] api/datastore: move s3 index upload helper to datastore backend Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 08/23] datastore: refactor chunk insert based on backend Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 09/23] verify: rename corrupted to corrupt in log output and function names Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 10/23] verify/datastore: make rename corrupt chunk a datastore helper method Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 11/23] datastore: refactor rename_corrupt_chunk error handling Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 12/23] chunk store: implement per-chunk file locking helper for s3 backend Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 13/23] datastore: acquire chunk store mutex lock when renaming corrupt chunk Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 14/23] datastore: get per-chunk file lock for chunk rename on s3 backend Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 15/23] fix #6961: datastore: verify: evict corrupt chunks from in-memory LRU cache Christian Ebner
2025-11-05 12:22 ` Christian Ebner [this message]
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 17/23] GC: fix race with chunk upload/insert on s3 backends Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 18/23] GC: lock chunk marker before cleanup in phase 3 " Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 19/23] datastore: GC: drop overly verbose info message during s3 chunk sweep Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 20/23] chunk store: reduce exposure of clear_chunk() to crate only Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 21/23] chunk store: make chunk removal a helper method of the chunk store Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 22/23] GC: fix deadlock for cache eviction and garbage collection Christian Ebner
2025-11-05 12:22 ` [pbs-devel] [PATCH proxmox-backup v3 23/23] chunk store: never fail when trying to remove missing chunk file Christian Ebner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20251105122233.439382-17-c.ebner@proxmox.com \
--to=c.ebner@proxmox.com \
--cc=pbs-devel@lists.proxmox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.