Merge pull request #50 from Aiven-Open/daniel.blasina/add_error_messages_to_exceptions

migration: adding error message to Exception explaining what happened

Author: vladislav-manafov-aiven

aiven_mysql_migrate/migration.py

@@ -198,7 +198,7 @@ def _check_database_size(self, max_size: float):
             )
             source_size = cur.fetchone()["size"] or 0
         if source_size > max_size:
-            raise DatabaseTooLargeException()
+            raise DatabaseTooLargeException(f"Database is larger than the maximum size of {max_size} bytes")
 
     def _check_bin_log_format(self):
         with self.source.cur() as cur:
@@ -215,7 +215,7 @@ def _check_source_target_uuids_aligned(self):
                 (source_server_uuid,)
             )
             if cur.fetchone()["included"] != 1:
-                raise IncompatibleGtidsException()
+                raise IncompatibleGtidsException("gtid_executed on target server doesn't match gtid from source")
 
     def run_checks(
         self,
@@ -361,22 +361,23 @@ def _ensure_target_replica_running(self, check_interval: float = 2.0, retries: i
                 cur.execute("SHOW SLAVE STATUS")
                 rows = cur.fetchall()
                 if not rows:
-                    raise ReplicaSetupException()
+                    raise ReplicaSetupException("SHOW SLAVE STATUS didn't return any rows")
 
                 try:
                     slave_status = next(
                         row for row in rows
                         if row["Master_Host"] == self.source.hostname and row["Master_Port"] == self.source.port
                     )
                 except StopIteration as e:
-                    raise ReplicaSetupException() from e
+                    raise ReplicaSetupException("Replication didn't start, Master info not available") from e
 
                 if slave_status["Slave_IO_Running"] == "Yes" and slave_status["Slave_SQL_Running"] == "Yes":
                     return
 
                 time.sleep(check_interval)
 
-            raise ReplicaSetupException()
+            raise ReplicaSetupException(f"Replication didn't start after {retries} "
+                                        f"retries with interval {check_interval} seconds")
 
     def _wait_for_replication(self, *, seconds_behind_master: int = 0, check_interval: float = 2.0):
         LOGGER.info("Wait for replication to catch up")
@@ -386,19 +387,19 @@ def _wait_for_replication(self, *, seconds_behind_master: int = 0, check_interva
                 cur.execute("SHOW SLAVE STATUS")
                 rows = cur.fetchall()
                 if not rows:
-                    raise ReplicaSetupException()
+                    raise ReplicaSetupException("SHOW SLAVE STATUS didn't return any rows")
 
                 try:
                     slave_status = next(
                         row for row in rows
                         if row["Master_Host"] == self.source.hostname and row["Master_Port"] == self.source.port
                     )
                 except StopIteration as e:
-                    raise ReplicaSetupException() from e
+                    raise ReplicaSetupException("Replication didn't catch up, Master info not available") from e
 
                 lag = slave_status["Seconds_Behind_Master"]
                 if lag is None:
-                    raise ReplicaSetupException()
+                    raise ReplicaSetupException("Replication didn't catch up, Seconds_Behind_Master is null")
 
                 LOGGER.info("Current replication lag: %s seconds", lag)
                 if lag <= seconds_behind_master: