lukaszraczylo 29d57857ff fix: prevent MCP server hanging by adding concurrency, timeouts, and context propagation (#45) ...

Root cause: synchronous MCP request processing combined with missing
context propagation to the embedding layer caused indefinite hangs when
ONNX inference was slow or the database was contended.

Changes:
- MCP server: dispatch each request in its own goroutine with semaphore
  (cap 10) and WaitGroup for clean shutdown drain
- Embedding: add context-aware mutex acquisition (acquireMutex) so
  callers can bail out instead of blocking forever on a stuck ONNX model
- Vector client: propagate context through getOrComputeEmbedding and
  replace bare RLock() calls with context-aware acquireRLockWithContext
- Worker handlers: add 15s request-scoped timeouts to all search/context
  handlers (handleSearchByPrompt, handleContextInject, handleFileContext,
  handleContextCount, handleGetObservations/Summaries/Prompts)
- Worker HTTP server: set WriteTimeout=60s (was 0); SSE endpoint extends
  deadline per-request via http.ResponseController

Fixes #45

2026-05-26 14:29:34 +01:00

..

client_test.go

mnemonic ralphised (#24)

2026-01-11 12:41:28 +00:00

client.go

fix: prevent MCP server hanging by adding concurrency, timeouts, and context propagation (#45)

2026-05-26 14:29:34 +01:00

helpers_test.go

feat(leann-phase2): implement hybrid vector storage and graph-based search (#20)

2026-01-07 22:03:59 +00:00

helpers.go

feat(leann-phase2): implement hybrid vector storage and graph-based search (#20)

2026-01-07 22:03:59 +00:00

sync_test.go

Further improvements to the coverage.

2025-12-19 02:17:29 +00:00

sync.go

Make things 'betterer' across the board (#23)

2026-01-11 01:51:20 +00:00