HANA Slow for EKKO->EKPO->EKET

[Former Member]

Hi All,

I am trying to pick up details from EKET along with EKPO and EKKO. All the tables are created as Column tables. 

number of records in EKKO-> 230,990 , EKPO->2,282,844  and EKET -> 4,645,738

In HANA Studio it takes about 2 seconds to fetch 1000 records. And I still feel its slow

Just want to Benchmark with experienced Hana users whether this is normal or is there anyway I can improve speed .

Thanks

[mathanponnucham]

Dear Keshan,

SAP HANA system performance cannot be generalised. If you want to find the reason behind the slow performance of SAP HANA system then I would recommend you to start with the performance tab in the administration prespective.

then enable the different traces available in the tab trace configuration based on your previous analysis to find the root cause of the performance problem

Rgds,

Mat.

[Former Member]

Hi Mathan,

Yes seems like this is a system issue. Is there any way i can test performance ? And is there a way of comparing my system with some one else (maybe with SAP ) and measure my performance.

[mathanponnucham]

Dear Keshan,

You cannot compare any two SAP systems setup in different environment ( hardware, network,, system layout, number of records, etc).

You may find some info regarding SAP BW on HANA performance benchmarks in the link http://www.saphana.com/docs/DOC-2233. But as I said above no two SAP environments are the same.

Rgds,

Mat.

[rama_shankar3]

How is the general performance of the HANA  system.  It appears that you might have a system issue. Have you created any other tables in HANA and how are those tables performing when you query them using joins?

What is the replication method that you are using? SLT or BODS?

[Former Member]

I am using BODS. I have migrated other tables and they are the same .

[Former Member]

Hi,

I would suggest you to go for a HASH Partitioning across the hosts for these tables.

Its quite effective and superb method for large tables.

Cheers,

Papil

[lbreddemann]

Ahem... how about finding out why the query is slow first and rush to solutions afterwards?

just sayin'

[lbreddemann]

Would you mind to share the SQL or the attribute view definition you used?

Also the HANA revision would be interesting here.

- Lars

[hendrik_brandes]

Hello Keshan,

the question is, how do you fetch your records and what do you want to select?

Just by putting all data into HANA, it is no solution.

I have seen some examples, where even an optimized ABAP-code could overtake a HANA-select if the query has not been optimized with good conditions and joins.

Kind regards,

Hendrik

[Former Member]

how long does it take to group by all the records on any column except the PO line number?