[Chore](pipeline) set PipelineFragmentContext::_timeout and adjust dump_pipeline_tasks infomation display (#36023)

## Proposed changes
pick from #35328

be/src/pipeline/pipeline_fragment_context.h

@@ -71,6 +71,8 @@ public:
 
     bool is_timeout(const VecDateTimeValue& now) const;
 
+    int timeout_second() const { return _timeout; }
+
     PipelinePtr add_pipeline();
 
     PipelinePtr add_pipeline(PipelinePtr parent, int idx = -1);

be/src/runtime/fragment_mgr.cpp

@@ -797,16 +797,24 @@ std::string FragmentMgr::dump_pipeline_tasks(int64_t duration) {
     size_t i = 0;
     {
         std::lock_guard<std::mutex> lock(_lock);
-        fmt::format_to(debug_string_buffer, "{} pipeline fragment contexts are still running!\n",
-                       _pipeline_map.size());
+        fmt::format_to(debug_string_buffer,
+                       "{} pipeline fragment contexts are still running! duration_limit={}\n",
+                       _pipeline_map.size(), duration);
+
+        timespec now;
+        clock_gettime(CLOCK_MONOTONIC, &now);
         for (auto& it : _pipeline_map) {
             auto elapsed = (t - it.second->create_time()) / 1000000000.0;
             if (elapsed < duration) {
                 // Only display tasks which has been running for more than {duration} seconds.
                 continue;
             }
-            fmt::format_to(debug_string_buffer, "No.{} (elapse time = {}s, InstanceId = {}) : {}\n",
-                           i, elapsed, print_id(it.first), it.second->debug_string());
+            auto timeout_second = it.second->timeout_second();
+            fmt::format_to(debug_string_buffer,
+                           "No.{} (elapse_second={}s, query_timeout_second={}s, instance_id="
+                           "{}) : {}\n",
+                           i, elapsed, timeout_second, print_id(it.first),
+                           it.second->debug_string());
             i++;
         }
     }