Skip to Content
0
Former Member
Jul 06, 2015 at 06:18 AM

Stack Trace on Standby ASE of WS pair and Thread going down on Rep Server

165 Views

The inserts on two text column tables are failing with a stack trace in ASE with (infected with signal 11 at Address (ptn_clr_hintpage+0x24)) and following error in the Replication server errorlog.

I. 2015/07/03 05:26:41. The DSI thread for database 'HRS1SYBPRD15.gotrade2' is started.

I. 2015/07/03 05:26:41. A grouped transaction of 7 individual transactions has failed in database 'HRS1SYBPRD15.gotrade2'. Each transaction in the group will be executed individually.

I. 2015/07/03 05:26:41. Message from server: Message: 5702, State 1, Severity 10 -- 'ASE is terminating this process.'.

H. 2015/07/03 05:26:42. THREAD FATAL ERROR #5049 DSI EXEC(133(1) HRS1SYBPRD15.gotrade2) - dsiqmint.c(3275)

The DSI thread for database 'HRS1SYBPRD15.gotrade2' is being shutdown. DSI received data server error #5702 which is mapped to STOP_REPLICATION. See logged data server errors for more information. The data server error was caused by output command #2 mapped from input command #3 of the failed transaction.

I. 2015/07/03 05:26:42. The DSI thread for database 'HRS1SYBPRD15.gotrade2' is shutdown.

We are skipping these for now. Whenever an update on those skipped transaction comes its fails with following error on replication server.

I. 2015/07/03 07:12:15. DSI received Replication Server error #5185 which is mapped to WARN by error action mapping.

W. 2015/07/03 07:12:15. WARNING #5185 DSI EXEC(133(1) HRS1SYBPRD15.gotrade2) - /dsiexec.c(4289)

Row count mismatch for the command executed on 'HRS1SYBPRD15.gotrade2'. The command impacted 0 rows but it should impact 1 rows.

I. 2015/07/03 07:12:15. Function 'rs_get_textptr' for 'GoMLTradeMessage' returned no rows.

I. 2015/07/03 07:12:15. A grouped transaction of 11 individual transactions has failed in database 'HRS1SYBPRD15.gotrade2'. Each transaction in the group will be executed individually.

I. 2015/07/03 07:12:15. DSI received Replication Server error #5185 which is mapped to WARN by error action mapping.

W. 2015/07/03 07:12:15. WARNING #5185 DSI EXEC(133(1) HRS1SYBPRD15.gotrade2) - /dsiexec.c(4289)

Row count mismatch for the command executed on 'HRS1SYBPRD15.gotrade2'. The command impacted 0 rows but it should impact 1 rows.

E. 2015/07/03 07:12:16. ERROR #5149 DSI EXEC(133(1) HRS1SYBPRD15.gotrade2) - /dsiexec.c(5287)

Function 'rs_get_textptr' for 'GoMLTradeMessage' returned no rows.

We are skipping these also. We tried to create an error class for 5702 and 5185 but these transactions dont get skipped and DSI keeps trying to apply them at standby without getting down or moving ahead. At that time we see these errors in the replication error log.

I. 2015/07/03 05:41:59. The DSI thread for database 'HRS1SYBPRD15.gotrade2' is started.

E. 2015/07/03 05:42:01. ERROR #1027 DSI EXEC(133(1) HRS1SYBPRD15.gotrade2) - /dsiutil.c(339)

Open Client Client-Library error: Error: 84083974, Severity 5 -- 'ct_results(): network packet layer: internal net library error: Net-Library operation terminated due to disconnect'.

E. 2015/07/03 05:42:01. ERROR #5097 DSI EXEC(133(1) HRS1SYBPRD15.gotrade2) - /dsiutil.c(342)

The ct-lib function 'ct_results' returns FAIL for database 'HRS1SYBPRD15.gotrade2'. The errors are retryable. The DSI thread will restart automatically. See ct-lib messages for more information.

The workaround provided by SAP was to reboot which is not helping us as we keep getting these stack traces even after reboot and threads go down.

The solution provided by SAP was to use sp_reptostandby without use_index will take around 24 to 30 hours to run on our databases and it cannot be implemented soon.

Can anyone provide us some pointers on how to resolve these issues without using sp_reptostandby solution.

Regards,

Rhishikesh.