on 2016 Apr 22 10:18 AM
Hi,
Our HANA node has restarted after we incorporated row based security on cost center field using analytic privilege with repository procedure. Inside repository procedure, based on SESSION_USER respective Cost Center will be fetched from a HANA DB table whish is having around 20 K entries.
This analytic privilege is assigned to a general role which in turn is available for all users.
This resulted in the below crash dump , pointing to procedure call.
Could any one please let me know the root cause for this issue.
[CRASH_SHORTINFO] exception short info: (2016-03-13 14:35:59 032 Local)
SIGNAL 6 (SIGABRT) caught, sender PID: 98902, PID: 98902, thread: 581942599[thr=122700]: JobWrk3984, value int: 0, ptr: 0x0000000000000000, time: 2016-03-13 14:35:59 000 Local
[CRASH_EXTINFO] extended exception info: (2016-03-13 14:35:59 033 Local)
----> Dump of siginfo contents <----
signal: 6(SIGABRT)
code: -6(SI_TKILL: signal send by tkill)
sender pid: 98902
own pid: 98902
sender uid: 1002
own uid: 1002
value(int): 0 (0x0)
value(ptr): 0x0000000000000000
[CRASH_CONTEXT] context info: (2016-03-13 14:35:59 033 Local)
----> Crashing context information <----
ContextStack at (0x00007fcd07f574e0)
stack: 7fcd07b59000-7fcd07f58fff, size 4194304
guard: 7fcd07b49000-7fcd07b58fff, size 65536
last known stack pointer cur: 0, max: 0
ctx addr: 0x00007eec564d7020, ctx link: 0x00007eec564d7020, ctx owner: 0x00007eec564d7020
ctx name: JobWrk3984, ctx type: thread, ctx id: 581942599
ctx command text:
{
SQL: CALL "_SYS_BIC"."XXXXXXX.general.app-fin-gl-pcsviz/SP_FIN_PCSVIZ_CC_MNGR_COST_CTR_AUTH"(?)
Thanks
Regards,
Jeswanth
Hi,
I went through few notes related to error signal:6(SIGABRT) .
Does this error (signal:6(SIGABRT) ) usually happens due to memory allocation issues ?
I went through SAP OSS note 2229604 and issue with memory allocation is specified as the root cause.
Does the same hold true in our scenario as well ?
Thanks
Regards,
Jeswanth
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Jeswant,
No, that was a database decision based on an error found.
Looking at section [CRASH_STACK] you'll see the stack of the thread that triggered the DB to abort.
I suggest you to read SAP Note 2288661 - How-To Investigate a SAP HANA Crashdump and verify if there's an explanation for that. If not, please create an SAP Support incident and provide your crashdump.
BRs,
Lucas de Oliveira
Hi Jeswanth,
I couldn't find anything related. If you have the chance to upgrade your box to SP11 please do so and retest the scenario. If not or it this persists, please open an SAP support incident providing the indexserver crash file and the details to reproduce the issue.
BRs,
Lucas de Oliveira
Hello,
What is the current revision you are running? There is known issues which were fixed in revision 94 that relate to your crash. See 2138515 - SAP HANA SPS 09 Database Revision 94
As Lucas already pointed out, upgrading to the newest Maintenance Revision would be wise.
I would recommend going to Revision 112.02.
Here are some useful notes also:
SAP HANA Database Maintenance Revision 112.02 of SAP HANA platform software Support Package Stack (SPS) 11 is now available externally for media download on SAP Service Marketplace.
Please check SAP Note 2300417.
If you upgrade from a lower SAP HANA SPS it is recommended to update ALL other components (Studio, Modeler, DB Clients, DBSL, SLT, DS, ...) to at least the minimal version of SAP HANA SPS 11.
Please check also the SAP HANA Platform SPS 11 Release Note 2227464
Regards,
Michael
Hi Michael,
Could you please let us know the reason for this crash.
We experienced this index server crash after assigning dynamic analytic privilege to all users.
This analytic privilege will call a procedure which looks up into an attribute view based on SESSION_USER to derive the cost center value assigned .
Please find attached procedure code.
If no costcenter is assigned then the procedure returns 'EMPTY' as a value.
This index server crash also happens sporadically and even we cannot reproduce this issue on demand.
Could you please let us know the reason behind the index server crash.
Thanks
Regards,
Jeswanth
User | Count |
---|---|
71 | |
10 | |
8 | |
7 | |
6 | |
6 | |
6 | |
6 | |
6 | |
5 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.