I am trying to Migrate my Postgresql RDS to Aurora Postgresql using DMS.

All my tasks are failing, and I am unable to understand what can be done to help the tasks progress.

The error from cloud watch logs is

2019-05-29T15:05:48 [TASK_MANAGER ]I: Start loading table 'public'.'my_ios_device_table' (Id = 642) by subtask 3. Start load timestamp 00058A081DCF0B40 (replicationtask_util.c:707)
2019-05-29T15:05:48 [SOURCE_UNLOAD ]I: Calculated batch used for UNLOAD size is 1000 rows per fetch. (postgres_endpoint_unload.c:178)
2019-05-29T15:05:48 [SOURCE_UNLOAD ]I: REPLICA IDENTITY information for table 'public'.'my_ios_device_table': Query status='Success' Type='DEFAULT' Description='Old values of the Primary Key columns (if any) will be captured.' (postgres_endpoint_unload.c:191)
2019-05-29T15:05:49 [SOURCE_UNLOAD ]I: Unload finished for table 'public'.'my_ios_device_table' (Id = 642). 1589 rows sent. (streamcomponent.c:3392)
2019-05-29T15:05:49 [TARGET_LOAD ]I: Load finished for table 'public'.'my_ios_device_table' (Id = 642). 1589 rows received. 0 rows skipped. Volume transfered 1558848. (streamcomponent.c:3663)
2019-05-29T15:05:50 [TARGET_LOAD ]E: Command failed to load data with exit error code 1, Command output: <truncated> [1020403] (csv_target.c:981)
2019-05-29T15:05:50 [TARGET_LOAD ]E: Failed to wait for previous run [1020403] (csv_target.c:1578)
2019-05-29T15:05:50 [TARGET_LOAD ]E: Failed to load data from csv file. [1020403] (odbc_endpoint_imp.c:5648)
2019-05-29T15:05:50 [TARGET_LOAD ]E: Handling End of table 'public'.'my_ios_device_table' loading failed by subtask 3 thread 1 [1020403] (endpointshell.c:2416)
2019-05-29T15:05:50 [TASK_MANAGER ]W: Table 'public'.'my_ios_device_table' (subtask 3 thread 1) is suspended (replicationtask.c:2356)

I have checked the Postgresql Roles, Privileges and even disabled triggers using the SET session_replication_role = 'replica';.

I am at my wits end as to why this is failing and how to rectify it.

I am migrating from RDS Postgres 9.5.16 to Aurora Postgres 10.7

Any help is appreciated. Thanks

DMS Task Details

  • Have you checked the logs on the RDS instance? I was having the same errors and the RDS instance log showed me that DMS converted my geometry to varchar 255 and the values were too big. – TurboGus Jul 24 '19 at 16:15
  • Yes - I did. We saw that some indices failed to migrate - Any combinations I tried didn't help. So we did the next best thing - use pg_dump with the migration and creating an Aurora from Postgresql snapshot. We gave up on using DMS, because of this. – gvatreya Aug 20 '19 at 9:51
  • @gvatreya, Did you find any solution for this issue apart from pg_dump, I also got the same. In my case, DB tables are very large and it's hard to migrate with pg_dump. – Rj_N May 13 at 7:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.