Database constraint preventing continued scanning of multichain

[TronBlack]

When scanning the multichain chain, a triplet of ids is added to a table triggering an database Integrity exception. ABE continues to retry getting the same exception.

Symptom: Data stops updating in the explorer. The multichain chain continues just fine.

[gidgreen]

Thanks. Can you say anything about what happened in the transaction which triggered this? Or what sorts of things you were doing in general on the chain?

[TronBlack]

We were running load tests and adding 3000+ Assets. The problem is not with multichain, but with the explorer db construction.

We have a "fix" for it that allows it to continue. It is essentially just handling the Integrity exception so it doesn't get stuck.

I added two lines in Mce/SqlAbstraction.py.

@@ -405,6 +405,8 @@ class SqlAbstraction(object):
     def _execute(sql, stmt, params):
         try:
             sql.cursor().execute(stmt, params)
+        except (sql.module.IntegrityError) as e:
+            print "Found Integrity Error - continuing..."        
         except (sql.module.OperationalError, sql.module.InternalError, sql.module.ProgrammingError) as e:
             if sql.in_transaction or not sql.auto_reconnect:
                 raise

[gidgreen]

Thanks. Do you have any more information about the integrity exception that was thrown? Foreign key? Unique key? And what SQL statement was executing at the time?

[bitcartel]

@TronBlack Can you print the entire exception object?