Re: Jetstress performance strange results...
From: Richard Roddy [MSFT] (rroddy_at_online.microsoft.com)
Date: 04/22/04
- Next message: Richard Roddy [MSFT]: "Re: Routing internal mail externally - E5.5"
- Previous message: Richard Roddy [MSFT]: "Re: Bounce message"
- In reply to: Jéjé: "Re: Jetstress performance strange results..."
- Next in thread: Jéjé: "Re: Jetstress performance strange results..."
- Reply: Jéjé: "Re: Jetstress performance strange results..."
- Messages sorted by: [ date ] [ thread ]
Date: Thu, 22 Apr 2004 11:31:22 GMT
My opinion of what to put on the RAID 1 drives if you add them would be the
Exchange transaction logs.
Thanks,
Richard Roddy
Microsoft Exchange Support
This posting is provided "AS IS" with no warranties, and confers no rights.
--------------------
>From: "Jéjé" <willgart@_A_hAotmail_A_.com>
>References: <#braySCJEHA.3996@TK2MSFTNGP10.phx.gbl>
<etDBxwhJEHA.528@tk2msftngp13.phx.gbl>
<e2VdVitJEHA.3624@cpmsftngxa10.phx.gbl>
>Subject: Re: Jetstress performance strange results...
>Date: Tue, 20 Apr 2004 10:40:40 -0400
>Lines: 211
>X-Priority: 3
>X-MSMail-Priority: Normal
>X-Newsreader: Microsoft Outlook Express 6.00.2800.1409
>X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
>Message-ID: <eRMtpVuJEHA.1764@TK2MSFTNGP12.phx.gbl>
>Newsgroups: microsoft.public.exchange.admin
>NNTP-Posting-Host: 207.253.5.68
>Path:
cpmsftngxa10.phx.gbl!TK2MSFTNGXA05.phx.gbl!TK2MSFTNGP08.phx.gbl!TK2MSFTNGP12
phx.gbl
>Xref: cpmsftngxa10.phx.gbl microsoft.public.exchange.admin:423004
>X-Tomcat-NG: microsoft.public.exchange.admin
>
>Well,
>
>its what I've conclude myself, I have to change my disk subsystem.
>but I'm suprised by the poor performance of a system which has 10 times
more
>power then my old server.
>
>unfortunatly, I can't change my HD configuration for the moment because my
>server is in production and I've a lot of applications abvailable on it.
>* MS CRM
>* Reporting Services
>* Exchange 2003
>* Sharepoint windows services
>* home made ASP.NET applications
>* Index server
>* Live communication server 2003
>* Active directory
>* File sharing
>
>Its a SBS 2003 server, so, some of these applications are installed "by
>default".
>has you can see, I push this server in its limits ;-)
>Only ISA Server is installed on another computer...
>
>I currently looking for a 64mb write cache module for my smartarray
>controller. I except a big performance improvement on my server with this
>addon.
>
>If I add 2 drives in RAID 1 (stripped) what is your recommandation for the
>files to be put on this new set of drives?
>
>Jerome.
>
>
>"Richard Roddy [MSFT]" <rroddy@online.microsoft.com> a écrit dans le
message
>de news:e2VdVitJEHA.3624@cpmsftngxa10.phx.gbl...
>> Jetstress shows whether your disk configuration is an issue on an
Exchange
>> server.
>>
>> In your case, I have to say that the reason the difference is as small as
>> it is is because you are using RAID 5 on Server A and RAID 0 on Server B.
>>
>> For the database operations being tested, RAID 0 is much more efficient
>> than RAID 5. Thus, even the older hardware runs fairly well in comparison
>> to the new hardware.
>>
>>
>> Microsoft makes the following recommendations for storage of Databases:
>>
>> As you plan your storage solution for these files, implement a solution
>> that ensures reliability; in other words, RAID-0 is not a recommended
>> option. After reliability, your storage solution is based on a choice
>> between optimizing performance (RAID-1) and optimizing capacity (RAID-5).
>> If possible, use RAID-1 (or RAID-0+1) for these files.
>>
>> You can store public folders on a RAID-5 array because data in public
>> folders is usually written once and read many times. RAID-5 provides
>> improved read performance.
>>
>>
>> And these recommendations regarding Transaction Logs:
>>
>> You can significantly improve the performance and fault tolerance of
>> Exchange servers by placing each set of transaction log files on a
>separate
>> drive. Because each storage group has its own set of transaction log
>files,
>> the number of dedicated transaction log drives for your server should
>equal
>> the number of planned storage groups. With a Storage Area Network
>solution,
>> you can select a product to easily partition the virtualized space into
>> separate virtual drives for storage groups and transaction log files. In
>> addition, because transaction log files are critical to the operation of
a
>> server, you should protect the drives against failure, ideally by
hardware
>> mirroring using RAID. A RAID-0+1 configuration (in which data is mirrored
>> and then striped) is recommended.
>>
>>
>> As far as opening the email with the large attachment goes, unless the
>item
>> is edited, it is not moved from the STM to the EDB. More info on this can
>> be found in the following article:
>>
>> 232323 Native Content Storage in Microsoft Exchange
>> http://support.microsoft.com/?id=232323
>>
>> Thanks,
>> Richard Roddy
>> Microsoft Exchange Support
>>
>> This posting is provided "AS IS" with no warranties, and confers no
>rights.
>>
>> --------------------
>> >From: "Vinod Bhatt [MSFT]" <vbhatt@online.microsoft.com>
>> >References: <#braySCJEHA.3996@TK2MSFTNGP10.phx.gbl>
>> >Subject: Re: Jetstress performance strange results...
>> >Date: Mon, 19 Apr 2004 10:39:56 -0400
>> >Lines: 77
>> >X-Priority: 3
>> >X-MSMail-Priority: Normal
>> >X-Newsreader: Microsoft Outlook Express 6.00.2800.1409
>> >X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
>> >Message-ID: <etDBxwhJEHA.528@tk2msftngp13.phx.gbl>
>> >Newsgroups:
>microsoft.public.exchange.admin,microsoft.public.exchange.setup
>> >NNTP-Posting-Host: 208.29.145.9
>> >Path:
>>
>cpmsftngxa10.phx.gbl!TK2MSFTNGXA05.phx.gbl!TK2MSFTNGP08.phx.gbl!tk2msftngp1
3
>> phx.gbl
>> >Xref: cpmsftngxa10.phx.gbl microsoft.public.exchange.setup:108615
>> microsoft.public.exchange.admin:422791
>> >X-Tomcat-NG: microsoft.public.exchange.admin
>> >
>> >Jerome,
>> >I guess you question is about the slow response that your clients are
>> >seeing. I would recommend that you get collect and analyse the perfmon
on
>> >your exchange server.
>> >
>> >Check the following counters for your disk performance. (read the
>jetstress
>> >doc for more details on counters)
>> >
>> >- Physical disk\Average Disk sec/read>> This should usually average
>around
>> >40ms (0.040) with max of 100ms
>> >- Physical disk\Average Disk sec/write>> This should usually average
>around
>> >40ms (0.040) with max of 100ms
>> >- Physical Disk\ current disk queue length >> This should average around
>> 1.5
>> >x number of spindel
>> >
>> >Hope this helpd.
>> >-vb
>> >--
>> >---------------
>> >This posting is provided "AS IS" with no warranties, and confers no
>rights.
>> >---------------
>> >
>> >
>> >"Jéjé" <willgart@BBBhotmailAAA.com> wrote in message
>> >news:%23braySCJEHA.3996@TK2MSFTNGP10.phx.gbl...
>> >> Hi,
>> >>
>> >> I've just test my server with the jetstress tool
>> >> because I found my server not so responsive.
>> >> I've also tested another and old server with the jetstress tool at the
>> >same
>> >> time.
>> >>
>> >> My configs:
>> >> Server A= Xeon 2.8ghz + 1gb Ram + 4 * 36 Gb 10krpm SCSI RAID 5 + SBS
>2003
>> >> Server B= P2 350mhz + 256 mb ram + 3 * 9 Gb RAID 0 (Stripped) (yes, a
>> very
>> >> old server) + Win 2003 standard
>> >>
>> >> Creating 200 000 records take 1200seconds on the server A and 1500sec
>on
>> >the
>> >> server B!!!!
>> >> So, for my tests, the Server B is only 20 to 25% slower then the A.
>> >> and its not what I expect.
>> >>
>> >> What can I do or test ?
>> >>
>> >> My server A is my production server. I've only 15 users on it, but my
>> >> private store is "huge" (avg of 100mb by user) because we exchange a
>lot
>> >of
>> >> documents.
>> >>
>> >> From a client side, my server appear unresponsive when I open a email
>> with
>> >a
>> >> big attachment (10mb for example). From what I've read, exchange 2003
>> >store
>> >> an email in the STM database, when this EMail is opened through
>Outllook,
>> >> Exchange copy the content from the STM to the EDB database. I presume
>the
>> >> copy of this information between the 2 files takes a lot of time and
>the
>> >> server appear unresponsive.
>> >>
>> >> any idea ?
>> >>
>> >> For a new server, I'm surprised by the porr performance.
>> >> Last year my configuration has been: Server B + SBS 4.5 (with Exchange
>> >5.5)
>> >> + RAID 5 config instead-of strip and this old server with the same
>> private
>> >> store size had the same performance. (or a little more performant,
>> because
>> >I
>> >> never or rarelly get the unresponsive message)
>> >>
>> >> Thanks for your help.
>> >>
>> >> Jerome.
>> >>
>> >>
>> >
>> >
>> >
>>
>
>
>
- Next message: Richard Roddy [MSFT]: "Re: Routing internal mail externally - E5.5"
- Previous message: Richard Roddy [MSFT]: "Re: Bounce message"
- In reply to: Jéjé: "Re: Jetstress performance strange results..."
- Next in thread: Jéjé: "Re: Jetstress performance strange results..."
- Reply: Jéjé: "Re: Jetstress performance strange results..."
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|