Page 1 of 1

teradata connector stage as source

Posted: Mon Mar 05, 2012 6:15 am
by srivalli_b
Hi,

I have worked in teradata connector stage as target.But I did not work as source. Now I need to use as source. For increase performence can I know the what properties I need to set.

arrry-size- 2000
record count-2000
access method- immediate
parallel synchronization- no

Re: teradata connector stage as source

Posted: Mon Mar 05, 2012 12:46 pm
by kwwilliams
You're not really going to see performance improvement by working on the array size and the commit transaction. You could change the access method to bulk and then set the min and max sessions to settings recommended by your Teradata admins.

Posted: Tue Mar 06, 2012 4:29 am
by srivalli_b
Hi,

Thanks your mail.
my job:

teradata connector stage -> transformer -> teradata connector stage

But I am getting below warnings and fatal error


warning:
SRC_OPP_AVL,0: *** glibc detected *** /opt/IBM/InformationServer/Server/PXEngine/bin/osh: double free or corruption (out): 0x000000000230c760 ***
warning:
SRC_OPP_AVL,0: ======= Backtrace: =========
warning:
SRC_OPP_AVL,0: /lib64/libc.so.6[0x3a650750c6]
warning:
SRC_OPP_AVL,0: /opt/IBM/InformationServer/Server/PXEngine/lib/pxbridge.so(_ZNSbItSt11char_traitsItESaItEE4_Rep10_M_destroyERKS1_+0x9)[0x7f612f9511ff]
warning:
SRC_OPP_AVL,0: /opt/IBM/InformationServer/Server/PXEngine/lib/pxbridge.so(_Z29PXBridge_isMultiLinkSupportedRK10APT_StringP18CC_LoadedConnector+0xf1)[0x7f612f910fbb]
warning:
SRC_OPP_AVL,0: /opt/IBM/InformationServer/Server/PXEngine/lib/pxbridge.so(_ZN10PXBridgeOp10runLocallyEv+0x51c)[0x7f612f922644]
warning:
SRC_OPP_AVL,0: /opt/IBM/InformationServer/Server/PXEngine/lib/liborchx86_64.so(_ZN15APT_OperatorRep3runEv+0x488)[0x7f6138421068]

fatal error :
SRC_OPP_AVL,0: Operator terminated abnormally: received signal SIGABRT


Please help me out

Posted: Tue Mar 06, 2012 4:30 am
by srivalli_b
after changeing to bulk mode I am getting above warnings and fatal error.

Posted: Tue Mar 06, 2012 5:42 am
by srivalli_b
I set the MALLOC_CHECK_=0 in job level and administrator side also.

Now above warnings are gone. But job is kepping on running.

Please help me

Posted: Tue Mar 06, 2012 8:43 am
by kwwilliams
What version of Teradata are you using and which Teradata Conector stage variant are you using? You can find the Teradata Connector variant when you open the stage.

Posted: Tue Mar 06, 2012 8:43 am
by srivalli_b
Hi,

Job is keep on running because of Table is locked. But I am getting below error.


Info:

SRC_OPP_AVL,0: Generated SELECT statement: SELECT AVL_UID,SEQUENCE,DIR_UID,RES_UID,USR_ID,LASTUPDATE,RES_LEVEL,RSDATE,RFDATE,CLH_UID,PALLOC_UID,SRC_SYS_ID,LOAD_ERROR,DATA_ORIGIN,CREATED_EW_DTTM,LASTUPD_EW_DTTM,BATCH_SID FROM EFBI_DEV1_STG_S.CSPRO_OPP_AVL

Fatal error: SRC_OPP_AVL,0: Operator terminated abnormally: received signal SIGSEGV

Posted: Tue Mar 06, 2012 10:47 am
by kwwilliams
What version of Teradata are you using and which Teradata Conector stage variant are you using? You can find the Teradata Connector variant when you open the stage. (asked earlier seems we were posting at the same time)

Posted: Wed Mar 07, 2012 3:52 am
by srivalli_b
thanks kwwilliams . I changed varient as 12.it is working.

But I have one issue.

my table is having 6,00,000 rows. with bulk option up to 420000 rows loaded properly. whenever I select more than 420000 it is not loaded.
it is throwing below error.

SRC_OPP_AVL,0: RDBMS code 2646: No more spool space in tdadmin. SQL statement: select AVL_UID,SEQUENCE,DIR_UID,RES_UID,USR_ID,LASTUPDATE,RES_LEVEL,RSDATE,RFDATE,CLH_UID,PALLOC_UID,SRC_SYS_ID,LOAD_ERROR,DATA_ORIGin ,CREATED_EW_DTTM,LASTUPD_EW_DTTM,BATCH_SID from EFBI_DEV1_STG_s.CSPRO_OPP_AVl sample 500000 (CC_TeraParallelTransporter::getRow, file CC_TeraParallelTransporter.cpp, line 1,616)

source properties:

access method:Bulk
arrysize 2000
record count 2000
isolation level default

sleep 0
tenacity 0
maxsessions-8
maxpartition sessions- 0
min sessions- 8
max buffer size-1000000

Posted: Wed Mar 07, 2012 7:59 am
by chulett
srivalli_b wrote:RDBMS code 2646: No more spool space in tdadmin.
Take that error to your DBA.