Add bootstrap_room validation to detect metadata corruption in PD disaggregation (#17430)

Co-authored-by: 继优 <jiyou.ljy@alibaba-inc.com>
Co-authored-by: Shangming Cai <csmthu@gmail.com>

python/sglang/srt/disaggregation/decode.py

@@ -722,7 +722,12 @@ class DecodeTransferQueue:
     def extend(self, decode_reqs: List[DecodeRequest]) -> None:
         self.queue.extend(decode_reqs)
 
-    def _commit_transfer_to_req(self, decode_req: DecodeRequest) -> None:
+    def _commit_transfer_to_req(self, decode_req: DecodeRequest) -> bool:
+        """
+        Returns:
+            True if the request should be removed from the queue (success or corruption)
+            False if metadata not ready yet (keep in queue for next poll)
+        """
         idx = decode_req.metadata_buffer_index
         (
             output_id,
@@ -734,8 +739,48 @@ class DecodeTransferQueue:
             output_topk_p,
             output_topk_index,
             output_hidden_states,
+            output_bootstrap_room,
         ) = self.metadata_buffers.get_buf(idx)
 
+        # Validate bootstrap_room to detect context corruption
+        actual_room = output_bootstrap_room[0].item()
+        expected_room = (
+            decode_req.req.bootstrap_room
+            if decode_req.req.bootstrap_room is not None
+            else 0
+        )
+
+        if decode_req.req.bootstrap_host == FAKE_BOOTSTRAP_HOST or (
+            decode_req.req.bootstrap_host is None
+            and self.scheduler.server_args.disaggregation_decode_enable_fake_auto
+        ):
+            # Warm up or fake transfer mode
+            pass
+        elif actual_room == 0:
+            # Case 1: Metadata not ready yet (actual_room == 0)
+            # Keep request in queue and wait for next poll
+            return False
+        elif actual_room != expected_room:
+            # Case 2: Real corruption detected (mismatch)
+            # Abort the request and remove from the queue
+            error_msg = (
+                f"Context corruption detected: Request {decode_req.req.rid} "
+                f"(bootstrap_room={expected_room}) received metadata from "
+                f"bootstrap_room={actual_room}. "
+                f"Metadata buffer index: {idx}. "
+                f"This indicates metadata buffer index collision."
+            )
+            logger.error(error_msg)
+            prepare_abort(
+                decode_req.req,
+                "Metadata corruption detected - bootstrap_room mismatch",
+                status_code=HTTPStatus.INTERNAL_SERVER_ERROR,
+            )
+            decode_req.kv_receiver.clear()
+            decode_req.kv_receiver = None
+            return True
+
+        # Case 3: Success - commit the transfer
         decode_req.req.output_ids.append(output_id[0].item())
         decode_req.req.cached_tokens = cached_tokens[0].item()
         if not self.spec_algorithm.is_none():
@@ -765,6 +810,7 @@ class DecodeTransferQueue:
             auto_next_anon=True,
         )
         decode_req.req.time_stats.wait_queue_entry_time = time.perf_counter()
+        return True
 
     def pop_transferred(self, rids_to_check: Optional[List[str]] = None) -> List[Req]:
         if not self.queue:
@@ -800,8 +846,20 @@ class DecodeTransferQueue:
                     self.scheduler.metrics_collector.increment_transfer_failed_reqs()
                 continue
             elif poll == KVPoll.Success:
-                self._commit_transfer_to_req(decode_req)
+                should_remove = self._commit_transfer_to_req(decode_req)
+                if should_remove:
                     indices_to_remove.add(i)
+                    # Check if request was aborted due to corruption
+                    if isinstance(decode_req.req.finished_reason, FINISH_ABORT):
+                        self.scheduler.stream_output(
+                            [decode_req.req], decode_req.req.return_logprob
+                        )
+                        release_kv_cache(
+                            decode_req.req, self.tree_cache, is_insert=False
+                        )
+                        if self.scheduler.enable_metrics:
+                            self.scheduler.metrics_collector.increment_transfer_failed_reqs()
+                    else:
                         transferred_reqs.append(decode_req.req)
             elif poll in [
                 KVPoll.Bootstrapping,

python/sglang/srt/disaggregation/utils.py

@@ -132,6 +132,10 @@ class MetadataBuffers:
             self.output_hidden_states = torch.zeros(
                 (size, hidden_size), dtype=hidden_states_dtype, device=device
             )
+            # Request validation: store bootstrap_room to detect metadata corruption
+            self.bootstrap_room = torch.zeros(
+                (size, 8), dtype=torch.int64, device=device
+            )
 
     def get_buf_infos(self):
         ptrs = [
@@ -144,6 +148,7 @@ class MetadataBuffers:
             self.output_topk_p.data_ptr(),
             self.output_topk_index.data_ptr(),
             self.output_hidden_states.data_ptr(),
+            self.bootstrap_room.data_ptr(),
         ]
         data_lens = [
             self.output_ids.nbytes,
@@ -155,6 +160,7 @@ class MetadataBuffers:
             self.output_topk_p.nbytes,
             self.output_topk_index.nbytes,
             self.output_hidden_states.nbytes,
+            self.bootstrap_room.nbytes,
         ]
         item_lens = [
             self.output_ids[0].nbytes,
@@ -166,6 +172,7 @@ class MetadataBuffers:
             self.output_topk_p[0].nbytes,
             self.output_topk_index[0].nbytes,
             self.output_hidden_states[0].nbytes,
+            self.bootstrap_room[0].nbytes,
         ]
         return ptrs, data_lens, item_lens
 
@@ -180,6 +187,7 @@ class MetadataBuffers:
             self.output_topk_p[idx],
             self.output_topk_index[idx],
             self.output_hidden_states[idx],
+            self.bootstrap_room[idx],
         )
 
     def set_buf(self, req: Req):
@@ -222,6 +230,10 @@ class MetadataBuffers:
             self.output_hidden_states[req.metadata_buffer_index].copy_(
                 req.hidden_states_tensor
             )
+        # Store bootstrap_room for validation on decode side
+        self.bootstrap_room[req.metadata_buffer_index, 0] = (
+            req.bootstrap_room if req.bootstrap_room is not None else 0
+        )
 
 
 #########################