cancel
Showing results for 
Search instead for 
Did you mean: 

BAD ALLOCATION SAPTS_GET_ALL_DATA_RW

Former Member
0 Kudos

Hi,

A huge treatment(update macro) is blocking every livecache sessions (

DB ANALYZER : UTK1 : user tasks is waiting for CPU in state 'vbegexcl'

UTK... user tasks is waiting for CPU in state 'command wai')

triggers 3 erros in sim_session (long runtime..) 'APTS_AGGMULT_TIMESERIES'

is there a way to isolate livecache session?

by limiting for a session, parameter like : data cache, oms_heap_limit?

thanks for your help.

chris

                1. #!!# DbpError exception in S

2010-09-16 16:17:28 385 ERR 51000 OBJECT APTS_AGGMULT_TIMESERIES ##########

  1. COM

2010-09-16 16:17:28 385 ERR 51000 OBJECT build 50012.002 CL 374315 (5.0_HOT) on

2010-09-16 16:17:28 385 ERR 51000 OBJECT Solaris X86_64

OMS_Context::GetObjFr

2010-09-16 16:17:28 385 ERR 51000 OBJECT omLiveCacheBase_1 (/SAP_DB/76_REL/sunx86

2010-09-16 16:17:28 385 ERR 51000 OBJECT 64/genopt/sys/wrk/incl/SAPDB/Oms/OMSCo

2010-09-16 16:17:28 385 ERR 51000 OBJECT ntext.hpp:44) n

2010-09-16 16:17:28 385 ERR 51000 OBJECT omsTerminate will be triggered:

2010-09-16 16:17:28 385 ERR 51000 OBJECT SAPTS_AGGMULT_TIMESERIES

2010-09-16 16:17:28 385 ERR 51000 OBJECT file: /sapmnt/patch_scratch/ux3102/5.0_H

2010-09-16 16:17:28 385 ERR 51000 OBJECT OT/sys/wrk/incl/COMBase/api_functor/cbs_

2010-09-16 16:17:28 385 ERR 51000 OBJECT api_obj_functor.h

2010-09-16 16:17:28 385 ERR 51000 OBJECT line: 69

2010-09-16 16:17:28 385 ERR 51000 OBJECT DbpError -102 errorKind = 0 in SAPTS_AG

2010-09-16 16:17:28 385 ERR 51000 OBJECT GMULT_TIMESERIES

, SQL statement cancell

2010-09-16 16:17:28 385 ERR 51000 OBJECT ed ERROR -1000102 in SAPTS_AGGMULT_TIMES

2010-09-16 16:17:28 385 ERR 51000 OBJECT ERIES (file: /sapmnt/patch_scratch/ux310

2010-09-16 16:17:28 385 ERR 51000 OBJECT 2/5.0_HOT/sys/wrk/incl/COMBase/api_funct

2010-09-16 16:17:28 385 ERR 51000 OBJECT or/cbs_api_obj_func

2010-09-16 16:17:28 385 ERR 51000 OBJECT <a href="lcapo_com_trace_01.txt#$2QIJ000

2010-09-16 16:17:28 385 ERR 51000 OBJECT 000gdmO06b$w0G">For details see lcapo_co

2010-09-16 16:17:28 385 ERR 51000 OBJECT m_trace_01.txt</a>

2010-09-16 16:17:28 339 ERR 51000 OBJECT SAPTS_AGGMULT_TIMESERIES

2010-09-16 16:17:28 339 ERR 51000 OBJECT file: /sapmnt/patch_scratch/ux3102/5.0_H

2010-09-16 16:17:28 339 ERR 51000 OBJECT OT/sys/wrk/incl/COMBase/api_functor/cbs_

2010-09-16 16:17:28 339 ERR 51000 OBJECT api_obj_functor.h

2010-09-16 16:17:28 339 ERR 51000 OBJECT line: 69

2010-09-16 16:17:28 339 ERR 51000 OBJECT

Accepted Solutions (1)

Accepted Solutions (1)

former_member229109
Active Contributor
0 Kudos

Hello Chris,

1.

The liveCache routine SAPTS_AGGMULT_TIMESERIES (resp. the workprocess

which has called it) seems to have been manually aborted or canceled due

unexpectedly long runtime. Please check the system log - /nsm21 and corresponding dev log of WP on your system.

DBberror -102 means:

The SQL statement currently executed was terminated by user or operator

intervention.

User Action:

No user action is required in interactive operation. Reset the

transaction with a ROLLBACK statement.

The error happens, for example, if a user e.g. use the "stop transaction" command in an application and then the error -102 is written to the current tracefile and to system error messages. (works as designed)

2.

u201CliveCache user tasks in state 'vbegexcl'u2026u201D ::

Vbegexcl(Enter Region):

Task waits to enter a protected region

It will be helpful to collect more information during the application run using x_cons tool.

Also we need to know the system SP stack, LCA build and liveCache version, the data selection during the application run

=> I recommend to create the SAP message to the component u201CBC-DB-LCAu201D

In general, please review the SAP notes u2013 1081508, 833216 and 936058, 1112046 and 957711.

Please use the SAP note 1111426 to check the configuration of the liveCache parameters.

Thank you & best regards, Natalia Khlopina

Former Member
0 Kudos

Hello Nathalia,

Thanks for your helpful reply.

I reviewed all the note given, and check Livecache parameter with dbanalyzer_InstanceParameterCheck.cfg :

W2 Recommended value for parameter 'LOG_QUEUE_COUNT' is 'MAXCPU' (8), current value is 1

W1 Recommended value for parameter 'MP_RGN_LOOP' is 10000, current value is - 1

W1 Recommended value for parameter '_IOPROCS_SWITCH' is 0, current value is 2

But our question is more general :

could it be possible that a single task don't block all other tasks (it seems that what is happened in our case)

(load balancing)

Note 936058

Within a user kernel thread, tasks operate in a cooperative manner. In other words, they decide when to suspend themselves and allow processing of the next task (see task switch). As a result of this behavior, a single task may block all other tasks in the same user kernel thread.

For optimum distribution of the tasks among the UKTs, the database system performs internal load balancing. In this way, the database system achieves more efficient resource utilization and short response times for the individual users, especially on computers using several processors.

Thanks for help.

Best Regards,

Chris

Answers (0)