Hello hana experts !
I nead to regylar load geodata in sap hana and store. Now it works in oracle.
The aim is to load in to table hana faster then oracle table.
in oracle(my local machine):
i have a table
create table UDO_TEST1
(
id NUMBER,
wlat NUMBER,
wlong NUMBER
);
insert table
insert 1000 records performed for 0.032 s.
in hana ( server )
i have column table
CREATE COLUMN TABLE "SCHEMA"."ROW_TABLE_TEST1" ("ID" INTEGER CS_INT,
"WLAT" DOUBLE CS_DOUBLE,
"WLONG" DOUBLE CS_DOUBLE) UNLOAD PRIORITY 5 AUTO MERGE PARTITION BY 'RANGE ID 1-500,501-1000,*'
insert into table
insert 1000 records performed for 0.579 s.🤯
if i have a row table insert performed for 0.160 s. (oracle 0,032 s.)
How do I get insert to the table hana faster than Oracle, maybe someone has a similar experience ? I nead to load data on regular basis.
1 Answer
-
Best AnswerPosted on Sep 19, 2016 at 09:14 AM
Viktor Portyankin wrote:
How do I get insert to the table hana faster than Oracle, maybe someone has a similar experience ? I nead to load data on regular basis.
Well, I'd start with a realistic test scenario.
Your loop insert in Oracle will very likely be optimised by PL/SQL into one big bulk insert.
SQLScript doesn't do this and ends up with 1000 single inserts.
That way you're not testing the insert speed but the ability to short circuit static loops.
Alternatively you could write a INSERT FROM SELECT where you create the tuples in the select and pass this in one INSERT statement.
Add a comment
Add a comment