Skip to content

xet upload: fix chunk cache desync #1771

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Sep 26, 2025
Merged

xet upload: fix chunk cache desync #1771

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d4f26ec
xet upload: fix chunk cache desync
coyotte508 Sep 26, 2025
c2dd130
better commetn
coyotte508 Sep 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions packages/hub/.gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,3 @@
xet-core-wasm-build
shard.bin
xorb.bin
*.bin
.debug
10 changes: 10 additions & 0 deletions packages/hub/src/utils/ChunkCache.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,16 @@ export class ChunkCache {
hmacs = new Set<string>(); // todo : remove old hmacs

addChunkToCache(hash: string, xorbIndex: number, chunkIndex: number, hmac: string | null): void {
if (this.map.has(hash)) {
// Happens when we receive an existing chunk from remote dedup info (eg duplicate chunk in shard? Or shards with same hmac key
// sharing chunks/xorbs)

// processing this chunk again would desync the cache, as `this.map.size` would not increase, as opposed to `this.index`

// Ideally we'd still process it to evict it later ("refresh it") but would need more complex handling, or stop using
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note, maybe a not-to-complex solution:

  • remove and readd the key in the map (to put the key at the end)
  • in the eviction code, instead of checking on map.size > MAX_SIZE, check on map.size > 1 && firstValue === this.index (or something approaching)

Would need unit tests and careful thinking

// the Uint16Array / Int32Array which are optimized for memory usage
return;
}
this.map.set(hash, this.index);
if (hmac !== null) {
this.hmacs.add(hmac);
Expand Down