on ‎2024 Jan 09 8:28 PM
Dear Colleagues,
I've got a customer migrating from a competitor database to HANA (2.00.071). One of their jobs is taking ~30% longer on HANA. During tracing, I see a significant difference in duration for the statements involved.
ST12 trace shows an average duration per execution twice as much as the numbers I see (duravg ) on HANA with sqltrace. (e.g. +400 microsecs vs 190 microsecs on HANA etc..)
How can this be explained, and what would you recommend the next step debugging this?
Cheers,
Lucjan
Request clarification before answering.
Hi,
ST12 (like SAT) measures in the ABAP VM while ST05 measures on database interface level (closer to the DB) and HANA traces measure directly on the DB.
So ST12 / SAT measurements contain more executed code in the stack than ST05 traces and HANA traces even less executed code and no network communication if measured directly on the DB server.
Probably not helpful anymore after 1 year... 😉
Kind regards,
Hermann
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
| User | Count |
|---|---|
| 9 | |
| 8 | |
| 5 | |
| 4 | |
| 4 | |
| 2 | |
| 2 | |
| 2 | |
| 2 | |
| 2 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.