Sybase NNTP forums - End Of Life (EOL)

The NNTP forums from Sybase - forums.sybase.com - are now closed.

All new questions should be directed to the appropriate forum at the SAP Community Network (SCN).

Individual products have links to the respective forums on SCN, or you can go to SCN and search for your product in the search box (upper right corner) to find your specific developer center.

New system performance (Sybase 15.5)

4 posts in General Discussion Last posting was on 2012-02-23 15:29:41.0Z
SteveS Posted on 2012-02-21 20:02:40.0Z
Sender: 24ad.4f43f23a.1804289383@sybase.com
From: SteveS
Newsgroups: sybase.public.ase.general
Subject: New system performance (Sybase 15.5)
X-Mailer: WebNews to Mail Gateway v1.1t
Message-ID: <4f43f860.2616.1681692777@sybase.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="-=_forums-1-dub4f43f860"
NNTP-Posting-Host: 172.20.134.41
X-Original-NNTP-Posting-Host: 172.20.134.41
Date: 21 Feb 2012 12:02:40 -0800
X-Trace: forums-1-dub 1329854560 172.20.134.41 (21 Feb 2012 12:02:40 -0800)
X-Original-Trace: 21 Feb 2012 12:02:40 -0800, 172.20.134.41
Lines: 82
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:30935
Article PK: 73823

We're moving From HPUX PARISC to HPUX Itanium. I have a new
server,a new EMC VNX5100 array and of course a new 15.5
install. We're currently running 15.0.3 in production. I've
seen some things that give me concerns that something isn't
right.(New hardware/OS or Sybase?). I've set up a simple
test to test system performance. I am doing a table scan of
a ~1GB table forcing a 2KB prefetch. I have set Sybase cache
to 100MB to eliminate most caching. I have disabled the
array cache to eliminate that variable as well. I have the
entire system to myself with the single task running. Does
anyone have any ideas about the vast differences in run/cpu
times per iteration in the attached data?

Select count(*) from invoice_lines (index invoice_lines prefetch 2)


Stats for 10 iterations:

Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=456 total=555170), physical reads: (regular=836 apf=520236
total=521072), apf IOs used=520228
Adaptive Server cpu time: 200 ms. Adaptive Server elapsed time: 299003 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=40 total=554754), physical reads: (regular=8 apf=515816
total=515824), apf IOs used=515816
Adaptive Server cpu time: 5800 ms. Adaptive Server elapsed time: 89410 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=48 total=554762), physical reads: (regular=8 apf=515857
total=515865), apf IOs used=515857
Adaptive Server cpu time: 100 ms. Adaptive Server elapsed time: 233206 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=52 total=554766), physical reads: (regular=8 apf=515947
total=515955), apf IOs used=515947
Adaptive Server cpu time: 15200 ms. Adaptive Server elapsed time: 236070 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=40 total=554754), physical reads: (regular=8 apf=515987
total=515995), apf IOs used=515987
Adaptive Server cpu time: 5600 ms. Adaptive Server elapsed time: 42196 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=35 total=554749), physical reads: (regular=8 apf=515952
total=515960), apf IOs used=515952
Adaptive Server cpu time: 100 ms. Adaptive Server elapsed time: 237583 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=36 total=554750), physical reads: (regular=8 apf=515942
total=515950), apf IOs used=515942
Adaptive Server cpu time: 5500 ms. Adaptive Server elapsed time: 84226 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=44 total=554758), physical reads: (regular=8 apf=515957
total=515965), apf IOs used=515957
Adaptive Server cpu time: 700 ms. Adaptive Server elapsed time: 234283 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=52 total=554766), physical reads: (regular=8 apf=515947
total=515955), apf IOs used=515947
Adaptive Server cpu time: 5000 ms. Adaptive Server elapsed time: 43946 ms.
Table: invoice_lines scan count 1, logical reads: (regular=554714 apf=48 total=554762), physical reads: (regular=8 apf=515947
total=515955), apf IOs used=515947
Adaptive Server cpu time: 200 ms. Adaptive Server elapsed time: 235840 ms.


Sherlock, Kevin [TeamSybase] Posted on 2012-02-21 21:00:28.0Z
From: "Sherlock, Kevin [TeamSybase]" <kevin.sherlock@teamsybase.com>
Newsgroups: sybase.public.ase.general
References: <4f43f860.2616.1681692777@sybase.com>
Subject: Re: New system performance (Sybase 15.5)
Lines: 39
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2900.5512
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.5512
X-RFC2646: Format=Flowed; Original
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4f4405ec$1@forums-1-dub>
Date: 21 Feb 2012 13:00:28 -0800
X-Trace: forums-1-dub 1329858028 10.22.241.152 (21 Feb 2012 13:00:28 -0800)
X-Original-Trace: 21 Feb 2012 13:00:28 -0800, vip152.sybase.com
X-Authenticated-User: teamsybase
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:30936
Article PK: 73826

So many things to have to look at.

You have an entirely new disk storage backend? Raid type, Raid group config
(or storage pool config), rotational latency, LUN striping size, SAS/NL-SAS
differences , etc etc.

You have an entirely new OS? HBA settings, driver limits,
filesystem/rawdevice settings/limits, .

You have an entirely new ASE? Server configuration differences (around io,
network, time slice, engines, apf settings, page size, , locking scheme of
table, etc).

I'd start from the top down on this list.

Most of the new VNX systems have some sort of FAST cache, or other flash
drive staging (outside of the normal cache). Also, storage pool
configuration is critical. Thick LUNS, and homogeneous LUNS on your VNX? .

It would be interesting to get the WAIT data from MDA tables during each of
these iterations as well to see if you are waiting on OS resources, or some
kind of ASE resource.

<SteveS> wrote in message news:4f43f860.2616.1681692777@sybase.com...
> We're moving From HPUX PARISC to HPUX Itanium. I have a new
> server,a new EMC VNX5100 array and of course a new 15.5
> install. We're currently running 15.0.3 in production. I've
> seen some things that give me concerns that something isn't
> right.(New hardware/OS or Sybase?). I've set up a simple
> test to test system performance. I am doing a table scan of
> a ~1GB table forcing a 2KB prefetch. I have set Sybase cache
> to 100MB to eliminate most caching. I have disabled the
> array cache to eliminate that variable as well. I have the
> entire system to myself with the single task running. Does
> anyone have any ideas about the vast differences in run/cpu
> times per iteration in the attached data?
>


Cory Sane [TeamSybase] Posted on 2012-02-22 07:15:33.0Z
From: "Cory Sane [TeamSybase]" <cory!=sane>
Newsgroups: sybase.public.ase.general
References: <4f43f860.2616.1681692777@sybase.com> <4f4405ec$1@forums-1-dub>
In-Reply-To: <4f4405ec$1@forums-1-dub>
Subject: Re: New system performance (Sybase 15.5)
Lines: 49
MIME-Version: 1.0
Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=response
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Windows Mail 6.0.6002.18197
X-MimeOLE: Produced By Microsoft MimeOLE V6.0.6002.18463
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4f449615$1@forums-1-dub>
Date: 21 Feb 2012 23:15:33 -0800
X-Trace: forums-1-dub 1329894933 10.22.241.152 (21 Feb 2012 23:15:33 -0800)
X-Original-Trace: 21 Feb 2012 23:15:33 -0800, vip152.sybase.com
X-Authenticated-User: TeamSybase
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:30939
Article PK: 73830

You may also want to start with a simpler test...
use unix dd command to read the devices in your testing...
This would eliminate the ASE factor.


--
Cory Sane
[TeamSybase]
Certified Sybase Associate DBA for ASE 15.0

"Sherlock, Kevin [TeamSybase]" <kevin.sherlock@teamsybase.com> wrote in message news:4f4405ec$1@forums-1-dub...
> So many things to have to look at.
>
> You have an entirely new disk storage backend? Raid type, Raid group config
> (or storage pool config), rotational latency, LUN striping size, SAS/NL-SAS
> differences , etc etc.
>
> You have an entirely new OS? HBA settings, driver limits,
> filesystem/rawdevice settings/limits, .
>
> You have an entirely new ASE? Server configuration differences (around io,
> network, time slice, engines, apf settings, page size, , locking scheme of
> table, etc).
>
> I'd start from the top down on this list.
>
> Most of the new VNX systems have some sort of FAST cache, or other flash
> drive staging (outside of the normal cache). Also, storage pool
> configuration is critical. Thick LUNS, and homogeneous LUNS on your VNX? .
>
> It would be interesting to get the WAIT data from MDA tables during each of
> these iterations as well to see if you are waiting on OS resources, or some
> kind of ASE resource.
>
> <SteveS> wrote in message news:4f43f860.2616.1681692777@sybase.com...
>> We're moving From HPUX PARISC to HPUX Itanium. I have a new
>> server,a new EMC VNX5100 array and of course a new 15.5
>> install. We're currently running 15.0.3 in production. I've
>> seen some things that give me concerns that something isn't
>> right.(New hardware/OS or Sybase?). I've set up a simple
>> test to test system performance. I am doing a table scan of
>> a ~1GB table forcing a 2KB prefetch. I have set Sybase cache
>> to 100MB to eliminate most caching. I have disabled the
>> array cache to eliminate that variable as well. I have the
>> entire system to myself with the single task running. Does
>> anyone have any ideas about the vast differences in run/cpu
>> times per iteration in the attached data?
>>
>
>


Jason L. Froebe [TeamSybase] Posted on 2012-02-23 15:29:41.0Z
From: "Jason L. Froebe [TeamSybase]" <jason@froebe.net>
Organization: TeamSybase
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:10.0.2) Gecko/20120216 Thunderbird/10.0.2
MIME-Version: 1.0
Newsgroups: sybase.public.ase.general
Subject: Re: New system performance (Sybase 15.5)
References: <4f43f860.2616.1681692777@sybase.com> <4f4405ec$1@forums-1-dub> <4f449615$1@forums-1-dub>
In-Reply-To: <4f449615$1@forums-1-dub>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4f465b65@forums-1-dub>
Date: 23 Feb 2012 07:29:41 -0800
X-Trace: forums-1-dub 1330010981 10.22.241.152 (23 Feb 2012 07:29:41 -0800)
X-Original-Trace: 23 Feb 2012 07:29:41 -0800, vip152.sybase.com
Lines: 21
X-Authenticated-User: TeamSybase
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:30945
Article PK: 73835


On 02/22/2012 01:15 AM, Cory Sane [TeamSybase] wrote:
> You may also want to start with a simpler test...
> use unix dd command to read the devices in your testing...
> This would eliminate the ASE factor.
>
>

Actually, I'd run Bonnie++ instead of dd for performance metrics on disk
subsystems.

http://hpux.connect.org.uk/hppd/hpux/Sysadmin/bonnie++-1.03e/

dd isn't a performance testing tool and the metrics it can display are
unreliable.


jason

--
To stay young requires unceasing cultivation of the ability to unlearn
old falsehoods. -- Lazarus Long, "Time Enough For Love"