Remove a deadlock during shutdown.

If queries are still running when the shutdown is initiated, they will
finish _during_ the shutdown. In that case, they might request chunk
eviction upon unpinning their pinned chunks. That might completely
fill the evict request queue _after_ draining it during storage
shutdown. If that ever happens (which is the case if there are _many_
queries still running during shutdown), the affected queries will be
stuck while keeping a fingerprint locked. The checkpointing can then
not process that fingerprint (or one that shares the same lock). And
then we are deadlocked.

storage/local/storage.go

@@ -467,20 +467,19 @@ func (s *memorySeriesStorage) handleEvictList() {
 				s.maybeEvict()
 			}
 		case <-s.evictStopping:
-			// Drain evictRequests to not let requesters hang.
+			// Drain evictRequests forever in a goroutine to not let
+			// requesters hang.
+			go func() {
 				for {
-				select {
-				case <-s.evictRequests:
-					// Do nothing.
-				default:
+					<-s.evictRequests
+				}
+			}()
 			ticker.Stop()
 			glog.Info("Chunk eviction stopped.")
 			close(s.evictStopped)
 			return
 		}
 	}
-		}
-	}
 }
 
 // maybeEvict is a local helper method. Must only be called by handleEvictList.