Add TPC-E as a workload #1
Comments
* base config for asynchronous clients * asynchronous clients for Oracle * Converted SQL Server build and driver to tdbc::odbc * Updated Transaction Counter to tdbc::odbc * Added SQL Server event driven scaling script * SQL Server event driven scaling options * Added SQL Server event driven scaling GUI Options (and fixed issues in Oracle ones) * Updated SQL Server build * Fixed typo in PostgreSQL GUI stored procedure options * Fixed with/without output suppresed typo in Db2 and PostgreSQL * Modified SQL Server TPC-H test to use tdbc::odbc * Corrected log to temp does not accept value from XML config * Fixed Issue #75 TPC-H cannot be executed for AWS RDS Oracle. Added system user to build script. * Updated metrics display to remove descriptive name so does not overrun buffer * Removed GUI only informational messages from driver script and set default asynchronous client config to 10 * Added event driven scaling feature to Db2 * Added PostgreSQL event driven scaling script * Added PostgreSQL GUI options for event driven scaling * Added MySQL Event driven scaling script * Updated MySQL event driven scaling GUI options * Event driven scaling for redis
|
Is TPC-E in the works or still in the planning phase? Based on the published benchmarks my sense is that Oracle and IBM will not be publishing any anytime soon. Does this inhibit the team from moving forward with TPC-E? |
|
For what it's worth, supporting TPC-E doesn't make sense based on feedback from benchmark experts from database vendor and hardware vendor teams. Only SQL Server has published results for the benchmark. Here is a link to a technical paper that outlines the issues with TPC-E - From A to E: Analyzing TPC’s OLTP Benchmarks. In essence, the TPC-E benchmark doesn't take full advantage of the CPU, Memory, and Disk resources as you scale the benchmark up. |
|
That’s good to know. I haven’t been able to get EGenDriver to work so I’ve
been looking for an easier path.
On Tue, Nov 10, 2020 at 20:26 Bill Ramos ***@***.***> wrote:
For what it's worth, supporting TPC-E doesn't make sense based on feedback
from benchmark experts from database vendor and hardware vendor teams. Only
SQL Server has published results for the benchmark. Here is a link to a
technical paper that outlines the issues with TPC-E - From A to E:
Analyzing TPC’s OLTP Benchmarks
<https://openproceedings.org/2013/conf/edbt/TozunPKJA13.pdfl>. In
essence, the TPC-E benchmark doesn't take full advantage of the CPU,
Memory, and Disk resources as you scale the benchmark up.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#1 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AIHM7ORLZRMSIIWKFCZY6XDSPHY35ANCNFSM4GIFS3BA>
.
--
Stephen Moon
(615) 920-2415
|
|
Many thanks Bill for the excellent insights and the link to the paper. Yes, currently TPC-E is not planned for consideration until both Oracle and Db2 have published fully audited TPC-E benchmarks. What is in planning is to add the Hybrid CH benchmark to HammerDB #123. This will not be immediate, firstly v4.0 will be released and additional updates following this with the CH benchmark planned for the next major release. |
|
Hi Steve, This test was "brutal" on all database engines. In 2016, I worked with Microsoft to port their SQL Server version to Oracle, DB2, SAP HANA, Sybase ASE, PostgreSQL, MySQL, SQL Server, Azure SQL Database, and MemSQL. The version we created supported Azure and AWS clouds. Jamie Reding was our product team sponsor at Microsoft. Maybe he can provide you with the data generator algorithm for the benchmark. I'd like to help in this adventure. |
* Add sslmode support for PostgreSQL database Allow pg_sslmode argument to be passed into the postgres driver. Keep default argument value to prefer to make it compatible with default postgres value (refer to https://www.postgresql.org/docs/13/libpq-ssl.html) pg_sslmode can be changed as follows: hammerdb>dbset db pg Database set to PostgreSQL hammerdb>print dict Dictionary Settings for PostgreSQL connection { pg_host = localhost pg_port = 5432 pg_sslmode = prefer } ... hammerdb>diset connection pg_sslmode disable Changed connection:pg_sslmode from prefer to disable for PostgreSQL hammerdb>print dict Dictionary Settings for PostgreSQL connection { pg_host = localhost pg_port = 5432 pg_sslmode = disable } ... All new code of the whole pull request, including one or several files that are either new files or modified ones, are contributed under the BSD-new license. I am contributing on behalf of my employer Amazon Web Services, Inc. * Updates for #274 add sslmode support for PostgreSQL database (#1) * Fix invalid sslmode value for Transaction Counter * Fix can't read sslmode value when using asynchronous scaling * Fix wrong number of args with sslmode for TPROC-H Refresh Function * Add sslmode to XML config for PostgreSQL connect pool * Add PostgreSQL sslmode option to GUI interface for TPROC-C/H and TX Counter * Fix sslmode for TPROC-C Asynch with Connect Pool Co-authored-by: Steve Shaw <[email protected]>
|
As discussed at TPC-OSS meeting 14th Nov. |
Add TPC-E as a workload
The text was updated successfully, but these errors were encountered: