fix bug: limit retries in starting state (ydb-platform#3398)

vporyadke · Apr 3, 2024 · 2631f5b · 2631f5b
1 parent d547c4c
commit 2631f5b
Showing 1 changed file with 8 additions and 0 deletions.
diff --git a/ydb/core/kqp/compute_actor/kqp_scan_fetcher_actor.cpp b/ydb/core/kqp/compute_actor/kqp_scan_fetcher_actor.cpp
@@ -154,6 +154,14 @@ void TKqpScanFetcherActor::HandleExecute(TEvKqpCompute::TEvScanError::TPtr& ev)
     }
 
     if (state->State == EShardState::Starting) {
+        ++TotalRetries;
+        if (TotalRetries >= MAX_TOTAL_SHARD_RETRIES) {
+            CA_LOG_E("TKqpScanFetcherActor: broken tablet for this request " << state->TabletId
+                << ", retries limit exceeded (" << state->TotalRetries << "/" << TotalRetries << ")");
+            SendGlobalFail(NDqProto::COMPUTE_STATE_FAILURE, YdbStatusToDqStatus(status), issues);
+            return PassAway();
+        }
+
         if (FindSchemeErrorInIssues(status, issues)) {
             return EnqueueResolveShard(state);
         }