"CALL "CHECK_TABLE_CONSISTENCY" ( ?, ?, ? )"

[former_member57954]

Dear all,

in _SYS_STATISTICS.HOST_LONG_RUNNING_STATEMENTS we have a lot of expensive entries from type "CALL "CHECK_TABLE_CONSISTENCY" ( ?, ?, ? )"

The time stamp don't match to DB13 "Check". The user is the R/3 User. So what event is triggering this call?

[michael_devine · Accepted solution]

Hi,

There is a correction to CHECK_TABLE_CONSISTENCY in note 2189594.

# 2189594 - SAP HANA SPS 09 Database Maintenance Revision 97.01

It limits the amount of threads consumed. It could be relevant. I dont see the same correction in SPS08 though.

Regards,

Michael

[former_member57954 · Accepted solution]

Hi Michael,

do you have experiences with the new maintenance Revision 97.01? We have in the current revision a lot of alerts in _SYS_STATISTICS.HOST_LONG_RUNNING_STATEMENTS for the CHECK_TABLE_CONSISTENCY. Is this related to the fix in 97.01?

[former_member57954 · Accepted solution]

We have now running the System on 97.02 and the following issue:

When I search for check_max_concurrency or table_consistency_check, I can't find these values in the Configuration tab. Do you have a link, where the Syntax is described?

Could not execute 'CALL CHECK_TABLE_CONSISTENCY ('CHECK', NULL, NULL)' in 17:22.638 minutes . [129]: transaction rolled back by an internal error: exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed ; $size$=262000; $name$=StringContainer; $type$=pool; $inuse_count$=55; $allocated_size$=10955408 exception 70006952: AttributeEngine: not enough memory exception 70006952: AttributeEngine: not enough memory exception 70006952: AttributeEngine: not enough memory exception 1000002: Allocation failed exception 1000002: Allocation failed exception 1000002: Allocation failed ; $size$=65536; $name$=itab; $type$=pool; $inuse_count$=1922250; $allocated_size$=320170475356 exception 1000002: Allocation failed ; $size$=65536; $name$=itab; $type$=pool; $inuse_count$=1921543; $allocated_size$=320121981460 exception 1000002: Allocation failed

[Former Member · Accepted solution]

Hi Rene,

Going by the error you seem to come across OOM situation while the Consistency check is running on the HANA system

We are on HANA 97.02 and have not seen any such errors yet during Consistency check

In fact the below SAP release notes for 97.02 confirms such OOM situation has been corrected, so this one could be a new issue that needs a correction

http://service.sap.com/sap/support/notes/2200738

Please check if any OOM dump has been generated and open a SAP support incident to understand the cause of the error

Hope it helps

Sunil

[michael_devine]

Hi,

This is not from a scheduled checkdb?

Regards,

Michael

[former_member57954]

Hi Michael, I assume not. CHECKDB was running on 2015-08-15, on 2015-08-08, on 2015-08-05 and 2015-08-03.

Messages in HOST_LONG_RUNNING_STATEMENTS are from: 08-16, 08-15, 08-11, 08-10 all over the day.

Could this related to the Backup Client?

[Former Member]

Hi Rene,

What is the size of the HANA System and what is the usual run time of the FULL consistency check on your system?

Such events may not have been triggered by a backup client it is an event that occurs when we have a long running consistency check process

Sunil

[former_member57954]

Hi Sunil, yep it seems it is simply from DBCHECK. I started yesterday a CHECK Job in ABAP and is running in the meantime fro 85.600 seconds and not ended.

We have a 4 Node Cluster á 2TB with about 1TB colum store on Node 2 and 3 and about 150 GB row store on Node 1. Node 4 ist the Stand By Node.

It theems it is a bit slowly, right? We are on SPS8 1.00.85.03.400544 (NewDB100_REL)