Re: HammerDB

From: Kellyn Pot'Vin-Gorman <dbakevlar_at_gmail.com>
Date: Tue, 11 Apr 2023 11:30:29 -0700
Message-ID: <CAN6wuX2e+98NA6UFcPwr45t7NFDEzxHdrKz=bMZxPuGKvvsB_Q_at_mail.gmail.com>





*Waves to Clay and Sallie

I think what gets me about HammerDB and Oracle is the last person to really invest in it is Kyle Hailey and I believe it was when he was back at DBArtisan. That's a looooong time ago and SLOB, FIO and others offer so much more for Oracle benchmarking. If you're SQL Server, it's more prominent- Mike Fal will swear by it, but I have a few peers in the SQL space that won't touch it with a 10 foot pole. It's still a go-to by many SQL Server folks, but TPM is not a clear benchmark and I was really surprised at how little understanding some of those using the tool had when running it against Oracle.

Nope- just go SLOB or FIO and do it right. Nothing against Swingbench or even Calibrate IO, (yes, we've been seeing this used for benchmarking, too) but we like to use what we know and what knows what we're working with.

Kellyn

On Tue, Apr 11, 2023 at 11:21 Clay Jackson <dmarc-noreply_at_freelists.org> wrote:

> Feeling compelled to jump in here to echo Kellyn’s comments from her blog,
> especially the introduction about why benchmarks are generally not as
> “useful” as they might appear.
>
>
>
> (Right)Sizing ANY platform (database/OS) for ANY application is not
> something to be undertaken lightly, as can be seen from many of the posts
> here on this list, and the many books and other material written about
> database performance. The authors of HammerDB already caution against
> using it to “compare platforms”; and using it to compare two databases as
> different as Oracle and SQLServer, especially when trying to size the
> database for an application that is NOT HammerDB just seems like a waste of
> time and a recipe for disappointment, hopefully after the OP has moved on.
>
>
>
> I think passing these suggestions “up the food chain” is the wisest course
> of action. Setting up a test environment to TEST YOUR application, or
> even the parts of it you already believe are “performance hogs” would be
> infinitely more revealing than some “random” benchmark.
>
>
>
> Clay Jackson
>
>
>
>
>
>
>
> *From:* oracle-l-bounce_at_freelists.org <oracle-l-bounce_at_freelists.org> *On
> Behalf Of *Sallie Cottingham
> *Sent:* Tuesday, April 11, 2023 11:03 AM
> *To:* Kellyn Pot'Vin-Gorman <dbakevlar_at_gmail.com>; mdemenko_at_gmail.com
> *Cc:* ORACLE-L <oracle-l_at_freelists.org>
> *Subject:* RE: HammerDB
>
>
>
> *CAUTION:* This email originated from outside of the organization. Do not
> follow guidance, click links, or open attachments unless you recognize the
> sender and know the content is safe.
>
>
>
> Hi Kellyn!
> I should’ve known that you would have and awesome blogpost on this
> subject. Very good info!
> The lead dev over the Oracle system who also has some experience with SQL
> is the person that requested we run this. The new SQL db will be on-prem
> and he says that it will be magic to ensure proper hardware sizing. I
> don’t necessarily believe this but was looking into it nonetheless. As you
> stated, increase in hardware/computing power doesn’t take care of poor SQL.
>
> I’ll be passing along your blog to my director.
>
>
>
> Thanks!
>
>
>
> *Sallie A. Cottingham*, OCP
>
> *Database Administrator*
>
> Comptroller of the Treasury
>
> Technology Solutions
>
> 425 Rep. John Lewis Way N. | Nashville, TN 37243
> <https://www.google.com/maps/search/425+Rep.+John+Lewis+Way+N.+%7C+Nashville,+TN+37243?entry=gmail&source=g>
>
> *sallie.cottingham_at_cot.tn.gov <sallie.cottingham_at_cot.tn.gov>* | Direct
> Line 615.401.7962
>
>
>
> [image: A picture containing food Description automatically generated]
>
>
>
> *From:* Kellyn Pot'Vin-Gorman <dbakevlar_at_gmail.com>
> *Sent:* Tuesday, April 11, 2023 12:01 PM
> *To:* mdemenko_at_gmail.com
> *Cc:* ORACLE-L <oracle-l_at_freelists.org>; Sallie Cottingham <
> Sallie.Cottingham_at_cot.tn.gov>
> *Subject:* Re: HammerDB
>
>
>
> I’m not a big fan of HammerDB with Oracle…
>
> https://dbakevlar.com/2021/04/hammerdb-cli-for-oracle-on-azure/
>
>
>
>
>
> On Tue, Apr 11, 2023 at 09:25 Maxim <mdemenko_at_gmail.com> wrote:
>
> Did you notice, there is
> https://github.com/TPC-Council/HammerDB/releases/download/v4.7/HammerDB-4.7-RHEL8.tar.gz
> - an extra tarball for RHEL8 ?
>
> On my rhel (oel) 8 box it starts without any errors (it is bundled with
> tcl/tk 8.6 compiled against glibc 28 - which is standard for rhel8)
>
>
>
> Regards
>
>
>
> Maxim
>
>
>
> On Tue, Apr 11, 2023 at 5:45 PM Sallie Cottingham <
> Sallie.Cottingham_at_cot.tn.gov> wrote:
>
> Hi –
>
> Anyone have experience using HammerDB for load testing an Oracle DB? We
> are looking to move an Oracle DB to MS SQL and want to run HammerDB on both
> to compare results.
>
> I’ve been using the hammerdb.com docs and GitHub discussions to get the
> software and install on my Linux server. The installation was fine but
> when I try to run it, I end up with errors talking about the version of
> GLIBC. (wish8.6: /lib64/libm.so.6: version `GLIBC_2.29' not found
> (required by ./lib/libtk8.6.so))
>
> I am a DBA and not a Linux admin so I’m hesitant to do anything that might
> affect the os (RHEL 8). The google search results warn against messing
> with GLIBC.
>
>
>
> If anyone has used HammerDB for an Oracle DB on RHEL Linux, I would love
> to chat with you!
>
> Thanks!
>
>
>
> *Sallie A. Cottingham*, OCP
>
> *Database Administrator*
>
> Comptroller of the Treasury
>
> Technology Solutions
>
> 425 Rep. John Lewis Way N. | Nashville, TN 37243
> <https://www.google.com/maps/search/425+Rep.+John+Lewis+Way+N.+%7C+Nashville,+TN+37243?entry=gmail&source=g>
>
> *sallie.cottingham_at_cot.tn.gov <sallie.cottingham_at_cot.tn.gov>* | Direct
> Line 615.401.7962
>
>
>
> [image: A picture containing food Description automatically generated]
>
>
>
> --
>
>
>
>
>
> *Kellyn Gorman*
>
> DBAKevlar Blog <http://dbakevlar.com/>
>
> about.me/dbakevlar
>
>
>
>
>



--
http://www.freelists.org/webpage/oracle-l


Received on Tue Apr 11 2023 - 20:30:29 CEST

Original text of this message