Issues

• A 2xx batch response that isn't valid JSON, or has a mistyped field, returns 500 instead of 502 (e.g. an SSO page served with 200)

  bug
  ktsu-dev/GitLfsCache#81 · Filter by author matt-edmondson opened on Sep 29, 2026
  1

• No container image has been published since 2026-08-26: the CI migration dropped the ghcr publish job, but README and deploy/k8s still pull ghcr.io/ktsu-dev/gitlfscache:latest

  bug
  ktsu-dev/GitLfsCache#80 · Filter by author matt-edmondson opened on Sep 29, 2026
  1

• locks/batch failure results drop upstream's lock object, so a 409 conflict no longer says who holds the file, and exhausted retries always report 429

  enhancement
  ktsu-dev/GitLfsCache#78 · Filter by author matt-edmondson opened on Sep 29, 2026
  1

• POST locks/batch returns 500 instead of 400 when ref is not an object (e.g. "ref": "refs/heads/main")

  bug
  ktsu-dev/GitLfsCache#77 · Filter by author matt-edmondson opened on Sep 29, 2026
  1

• No bytes-evicted metric: eviction sweeps are only visible in logs (spec instrumentation item missing)

  enhancement
  ktsu-dev/GitLfsCache#75 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• One network error on a single locks/batch item turns the whole fan-out into a 500, discarding locks upstream already granted

  bug
  ktsu-dev/GitLfsCache#74 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• A backslash in the request path escapes the upstream base-URL prefix and the repository allow-list

  bug
  ktsu-dev/GitLfsCache#73 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• No lock snapshot-size metric: operators can't see how close a repository is to MaxSnapshotLocks (spec instrumentation item missing)

  enhancement
  ktsu-dev/GitLfsCache#72 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• Unlock by path fails with 404/409 when the snapshot names a released-and-reissued lock id: the spec's one refresh + one retry isn't implemented

  enhancement
  ktsu-dev/GitLfsCache#71 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• A lock created or released during an in-flight lock-list refresh is missing from listings for a full ListTtl: the refresh publishes its stale walk after the invalidation ran

  bug
  ktsu-dev/GitLfsCache#70 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• A staging flush that fails (e.g. disk full) turns an upload upstream accepted into a 500, and leaves the staging file uncollectable until restart

  bug
  ktsu-dev/GitLfsCache#69 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• Batch responses advertise the proxy's 1h token lifetime even when upstream's href expires sooner, so git-lfs never re-batches and gets 401/403

  ktsu-dev/GitLfsCache#66 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• A repository with more locks than MaxSnapshotLocks is re-walked up to the ceiling on every GET locks, then relayed anyway

  ktsu-dev/GitLfsCache#65 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• A refused lock-list refresh returns a bare status to the client, dropping upstream's body and Retry-After / LFS-Authenticate headers

  ktsu-dev/GitLfsCache#64 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• Unlocking many paths through locks/batch can run a full lock-list walk per path, outside single-flight and the upstream limiter

  ktsu-dev/GitLfsCache#63 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• A Range request for a cached object returns the whole object with 200, so resumed git-lfs downloads start over

  ktsu-dev/GitLfsCache#62 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• When the leader request is aborted, every waiting follower fetches the same object from upstream separately (N+1 fetches instead of one)

  ready
  ktsu-dev/GitLfsCache#58 · Filter by author matt-edmondson opened on Sep 28, 2026
  1

• Every object download and upload returns 500 when the upstream key contains a dot (e.g. gitlab.com), though startup validation and batch calls accept it

  ktsu-dev/GitLfsCache#56 · Filter by author matt-edmondson opened on Sep 28, 2026
  2

• Lock-list snapshots are never evicted: every distinct (repository, refspec) keeps a full lock list in memory for the life of the process

  ready
  ktsu-dev/GitLfsCache#50 · Filter by author matt-edmondson opened on Sep 27, 2026
  1

• Upstream key casing splits the object cache: the same object via /GitHub/ and /github/ is fetched and stored twice

  ready
  ktsu-dev/GitLfsCache#49 · Filter by author matt-edmondson opened on Sep 27, 2026
  1

• Delegate repository-path glob matching to ktsu.TextFilter

  ktsu-reuse
  ktsu-dev/GitLfsCache#33 · Filter by author matt-edmondson opened on Sep 15, 2026
  6