Re: [BUGS] BUG #14530: Logical Decoding Slowness

I might have given you the create table perf data for the subtran case. 

We profiled again with debug info on today, and the results are

Subtran:

  Children      Self  Command Shared Object                       Symbol
+   94.84%     0.00%  postgres postgres                            [.] ExecMakeTableFunctionResult
+   94.84%     0.00%  postgres postgres                            [.] pg_logical_slot_get_changes_guts
+   94.80%     0.00%  postgres postgres                            [.] LogicalDecodingProcessRecord
+   94.45%     0.03%  postgres postgres                            [.] ReorderBufferCommit
+   94.19%     0.01%  postgres postgres                            [.] ReorderBufferCleanupTXN
+   94.11%     0.01%  postgres postgres                            [.] ReorderBufferReturnChange
+   94.08%    93.40%  postgres postgres                            [.] AllocSetFree

DDL:

+   93.88%     0.00%  postgres postgres                      [.] ExecMakeTableFunctionResult
+   93.88%     0.00%  postgres postgres                      [.] pg_logical_slot_get_changes_guts
+   93.83%     0.01%  postgres postgres                      [.] LogicalDecodingProcessRecord
+   93.78%     1.10%  postgres postgres                      [.] ReorderBufferCommit
+   87.69%     3.12%  postgres postgres                      [.] LocalExecuteInvalidationMessage
+   46.32%    46.21%  postgres postgres                      [.] CatalogCacheIdInvalidate
+   37.17%     0.37%  postgres postgres                      [.] RelfilenodeMapInvalidateCallback
+   36.71%    36.46%  postgres postgres                      [.] hash_seq_search
+    4.92%     4.91%  postgres postgres                      [.] CallSyscacheCallbacks