SAP Hana - SQL error

[joo_migueldimas]

Hello,

We have a problem/error in our SAP Hana database. When I open the Hana Studio our HQ1 system show the following error:

SAP DBTech JDBC: [2048]: column store error: search table error: [2613] executor: communication problem (com.sap.db.jdbc.exceptions.JDBCDriverException: SAP DBTech JDBC: [2048]: column store error: search table error: [2613] executor: communication problem)

In the SAP system, in ST22, I see several shortdumps:

For example the first one (CX_SY_OPEN_SQL_DB - SAPLSDIFRUNTIME):

SQL error "SQL code: 3584" occurred while accessing table "DDFTX".

Database error text: "SQL message: distributed SQL error:  [2617] executor:

plan operation execution failed with an exception"

I believe this two situations are connected!!

This hana server has two databases, the HD1 and HQ1, this server/appliance has 256Gb of RAM memory.

Can you help me please to solve this error?

Best regards,

JD

[Former Member · Accepted solution]

Hello JD,

What revision are you on ?

Are you using the ESS (Embedded Statistics Server) ?

If you are not using ESS then it could be possible that your statistics server has run out of memory and you will need to increase the allocationlimit.

Can you please upload the trace files.

KR,

Amerjit

[joo_migueldimas · Accepted solution]

Hello Amerjit,

This is Hana Platform Ed. 1.0 SPS08 (SAP HANA DB 1.00.82).

I searched information about ESS and I found the following video:

SAP HANA Academy - Monitoring - Embedded Statistics Server [SPS 09] - YouTube

In my Hana Studio I see in Landscape tab in Services the following information:

As you see in the previous image there is a statisticsserver service.

The statisticsserver has the value False in this hana database:

Do you recommend the activation of ESS? Can you explain me the advantages of that?

Thank you,

JD

[Former Member · Accepted solution]

Hello JD,

So you are still running with a dedicated statisticsserver. It's been recommended for quite some time now to migrate to ESS.

The briefest and simplest answer to the question would be that SAP simplified the HANA architecture (one less service to manage resulting is less memory usage and management footprint).

The following notes will help you through the steps to migrate to ESS or increase the allocationlimit of the statisticsserver.

1929538 - HANA Statistics Server - Out of Memory

1917938 - Migration of the statistics server for Revision 74 or higher

2092033 - Embedded Statistics Service Migration Guide

1925684 - ABAP adjustments for the new Embedded Statistic Server

Additionally, please go through the checks from the following note.

1999993 - SAP HANA Mini checks

Kind Regards,

Amerjit

[joo_migueldimas · Accepted solution]

Hello Amerjit,

We done the activation of ESS service and now everything is fine.

Until this moment we no longer have errors therefore I believe the activation of ESS solved our problem.

Thank you Amerjit,

Best regards,

João Dimas

[davidebruno]

Did you try to check into trace of index server?

it seems that your HANA database has same problem

[joo_migueldimas]

Hello Davide Bruno,

In the indexserver_alert_satdcha02.trc file I found the following lines:

...

pop.setLocale('BINARY')

[43067]{300848}[39/-1] 2015-12-06 19:49:07.482745 e Memory           ReportMemoryProblems.cpp(00694) : Printing information after out of memory occured. More information in after_oom_cleanup runtime dump!

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334159 e Memory           ReportMemoryProblems.cpp(00703) : Finished writing after_oom_cleanup runtime dump!

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334270 e join_eval        Estimator.cpp(00345) : exec estimation failed rc=1000002 error=1000002

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334325 e join_eval        MGetEstimations.cpp(00315) : estimateParallel failed rc=1000002

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334332 e join_eval        JoinEvaluator.cpp(00521) : , MGetEstimations failed for query part '0' with rc=1000002

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334366 e SERVER_TRACE     TRexApiSearch.cpp(10151) : IndexID: SYSTEM:0x7fe685ebf000:3:0:#432345564227580742 (432345564227580743):

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334472 e SERVER_TRACE     TRexApiSearch.cpp(10152) :     query: _SYS_STATISTICS:STATISTICS_ALERT_LAST_CHECK_INFORMATIONen/ALERT_ID (IN):'0'(EQ) _SYS_STATISTICS:STATISTICS_ALERT_LAST_CHECK_INFORMATIONen/ALERT_ID (IN):'0'(NE) OR

[43067]{300848}[39/-1] 2015-12-06 19:49:08.334495 e SERVER_TRACE     TRexApiSearch.cpp(10162) :     req.attr.: SNAPSHOT_ID,ALERT_ID,ALERT_LAST_CHECK_TIMESTAMP

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335382 e Executor         PlanExecutor.cpp(00690) : plan plan53867@satdcha02:30103 failed with rc 2048; column store errorsearch table error:  [1000002] message not found

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335434 e Executor         PlanExecutor.cpp(00690) : -- returns for plan53867@satdcha02:30103:

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335442 e Executor         PlanExecutor.cpp(00690) : search table error:  [1000002] message not found(2048), plan: 4 pops: RowPlanOperator pop1(out ab), RowPlanOperator pop2(in a, out c), RowPlanOperator pop3(in b, out d), RowPlanOperator pop4(in cd, out ef)

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335470 e Executor         PlanExecutor.cpp(00690) : pop1, 19:49:06.642  +1.692, cpu 140671033344904, RowPlanOperator, rc 2048, column store error

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335486 e Executor         PlanExecutor.cpp(00690) : Comm total: 0.000

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335489 e Executor         PlanExecutor.cpp(00690) : Total: 19:49:06.642 to 19:49:08.335 +1.693, cpu 140671033344904

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335501 e Executor         PlanExecutor.cpp(00690) : sizes a 0 b 0 f 0

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335505 e Executor         PlanExecutor.cpp(00690) : -- end executor returns

[43677]{300848}[39/-1] 2015-12-06 19:49:08.335506 e Executor         PlanExecutor.cpp(00690) : pop1 (rc 2048, column store error)

pop = RowPlanOperator() # pop1

pop.setDetails('')

[43677]{300848}[39/-1] 2015-12-06 19:49:10.026799 e TrexNet          BufferedIO.cpp(04284) : active channel 585 from 60456 to 127.0.0.1:30105: protocol mismatch - trying to read type Int but EOF reached

[43677]{300848}[39/-1] 2015-12-06 19:49:10.026838 e TrexNetIO        BufferedIO.cpp(01395) : dump of TrexNet buffer: size 29, current position 0x00007ff6567ae091 (29 bytes), mode incoming, channel 585, local port 60456, peer 127.0.0.1:30105, content:

BLOB START (Addr: 0x00007ff6567ae074 Len: 29)

BLOB END

1762897[thr=43677]: SqlExecutor at

1: 0x00007ff88bd647d4 in Execution::ContextFunctions::dumpInfo(Execution::Context&, ltt::basic_ostream<char, ltt::char_traits<char> >&, bool, bool, bool, bool, bool)+0x2c0 at ContextFunctions.cpp:629 (libhdbbasis.so)

2: 0x00007ff896837f45 in TrexNet::BufferedNetInput::checkType(unsigned char)+0x381 at BufferedIO.cpp:4286 (libhdbbasement.so)

3: 0x00007ff896838d80 in TrexNet::BufferedNetInput::operator>>(int&)+0x20 at BufferedIO.cpp:3557 (libhdbbasement.so)

...

Any tips?

BR,

JD