Quantcast

iometer on Linux Raw Disks

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

iometer on Linux Raw Disks

Prasad Kularatne

Hi folks,

I am planning to do some performance research on iSCSI disks mounted on RHEL 3.0 Update5.

I am planning to do this test on raw disks i.e. linux partitions.

I am using iometer 2004.07.30 version. When I attach dynamo to the iometer it returns the following message,

...Worker 5 running Access Spec: 32K; 75% Read; 0% random
Access specifications for test defined.
Grunt: Growing grunt data buffer from 32768 to 32768
Worker 0 setting targets...
Fail to get size for /dev/sdb1
Fail to get size for /dev/sdb1
Fail to get size for /dev/sdb1
Worker 1 setting targets...
Fail to get size for /dev/sdc1
Fail to get size for /dev/sdc1
Fail to get size for /dev/sdc1
Worker 2 setting targets...
Fail to get size for /dev/sdd1
Fail to get size for /dev/sdd1
Fail to get size for /dev/sdd1..."

But on the iometer I can select them and assign these partitions as raw disks to the workers. When I start the test, it shows as if it is running, results display changes indicating very large figures (e.g. 750MB/s/ 22,895 IOPs) etc. which cannot any way be realistic figures. After the specified time interval it stops, csv file indicates very large, unrealistic statistics.

But during the run no I/O is actually happening as evident by iostat o/p.

I am puzzled. Any ideas as to where I have done wrong?


Good Luck!
Prasad

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: iometer on Linux Raw Disks

Ming Zhang
run 07.30-ds2 version.

iometer might have problem on rhel3, but i do not have rhel3 to do test.

ming


On Fri, 2005-09-23 at 12:26 +0000, Prasad Kularatne wrote:

>
> Hi folks,
>
> I am planning to do some performance research on iSCSI disks mounted
> on RHEL 3.0 Update5.  
>
> I am planning to do this test on raw disks i.e. linux partitions.
>
> I am using iometer 2004.07.30 version. When I attach dynamo to the
> iometer it returns the following message,
>
> ...Worker 5 running Access Spec: 32K; 75% Read; 0% random
> Access specifications for test defined.
> Grunt: Growing grunt data buffer from 32768 to 32768
> Worker 0 setting targets...
> Fail to get size for /dev/sdb1
> Fail to get size for /dev/sdb1
> Fail to get size for /dev/sdb1
> Worker 1 setting targets...
> Fail to get size for /dev/sdc1
> Fail to get size for /dev/sdc1
> Fail to get size for /dev/sdc1
> Worker 2 setting targets...
> Fail to get size for /dev/sdd1
> Fail to get size for /dev/sdd1
> Fail to get size for /dev/sdd1..."
>
> But on the iometer I can select them and assign these partitions as
> raw disks to the workers. When I start the test, it shows as if it is
> running, results display changes indicating very large figures (e.g.
> 750MB/s/ 22,895 IOPs) etc. which cannot any way be realistic figures.
> After the specified time interval it stops, csv file indicates very
> large, unrealistic statistics.
>
> But during the run no I/O is actually happening as evident by iostat
> o/p.
>
> I am puzzled. Any ideas as to where I have done wrong?
>
>
> Good Luck!
> Prasad
>



-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server.
Download it for free - -and be entered to win a 42" plasma tv or your very
own Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: iometer on Linux Raw Disks

Prasad Kularatne

Hi Ming,

Thanks, the DS2 version works fine.

Good Luck!
Prasad

K K I P Kularatne
IT Specialist
Team Leader - SSPS
Integrated Technology Services
IBM Sri Lanka
Phone: 94 11 2493546
Fax: 94 11 2434594
Internet Mail: [hidden email]



Ming Zhang <[hidden email]>

09/23/2005 01:29 PM

Please respond to
mingz

To
Prasad Kularatne/Sri Lanka/IBM@IBMSG
cc
iometer-user-list <[hidden email]>
Subject
Re: [Iometer-user] iometer on Linux Raw Disks





run 07.30-ds2 version.

iometer might have problem on rhel3, but i do not have rhel3 to do test.

ming


On Fri, 2005-09-23 at 12:26 +0000, Prasad Kularatne wrote:
>
> Hi folks,
>
> I am planning to do some performance research on iSCSI disks mounted
> on RHEL 3.0 Update5.  
>
> I am planning to do this test on raw disks i.e. linux partitions.
>
> I am using iometer 2004.07.30 version. When I attach dynamo to the
> iometer it returns the following message,
>
> ...Worker 5 running Access Spec: 32K; 75% Read; 0% random
> Access specifications for test defined.
> Grunt: Growing grunt data buffer from 32768 to 32768
> Worker 0 setting targets...
> Fail to get size for /dev/sdb1
> Fail to get size for /dev/sdb1
> Fail to get size for /dev/sdb1
> Worker 1 setting targets...
> Fail to get size for /dev/sdc1
> Fail to get size for /dev/sdc1
> Fail to get size for /dev/sdc1
> Worker 2 setting targets...
> Fail to get size for /dev/sdd1
> Fail to get size for /dev/sdd1
> Fail to get size for /dev/sdd1..."
>
> But on the iometer I can select them and assign these partitions as
> raw disks to the workers. When I start the test, it shows as if it is
> running, results display changes indicating very large figures (e.g.
> 750MB/s/ 22,895 IOPs) etc. which cannot any way be realistic figures.
> After the specified time interval it stops, csv file indicates very
> large, unrealistic statistics.
>
> But during the run no I/O is actually happening as evident by iostat
> o/p.
>
> I am puzzled. Any ideas as to where I have done wrong?
>
>
> Good Luck!
> Prasad
>


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: iometer on Linux Raw Disks

Ming Zhang
cool.

Ming

On Sun, 2005-09-25 at 03:50 +0000, Prasad Kularatne wrote:

>
> Hi Ming,
>
> Thanks, the DS2 version works fine.
>
> Good Luck!
> Prasad
>
> K K I P Kularatne
> IT Specialist
> Team Leader - SSPS
> Integrated Technology Services
> IBM Sri Lanka
> Phone: 94 11 2493546
> Fax: 94 11 2434594
> Internet Mail: [hidden email]
>
>
> Ming Zhang <[hidden email]>
>
> 09/23/2005 01:29 PM
>          Please respond to
>                mingz
>
>
>
>
>                To
> Prasad
> Kularatne/Sri
> Lanka/IBM@IBMSG
>                cc
> iometer-user-list
> <iometer-
> [hidden email]>
>           Subject
> Re: [Iometer-
> user] iometer on
> Linux Raw Disks
>
>
>
>
>
>
>
>
> run 07.30-ds2 version.
>
> iometer might have problem on rhel3, but i do not have rhel3 to do
> test.
>
> ming
>
>
> On Fri, 2005-09-23 at 12:26 +0000, Prasad Kularatne wrote:
> >
> > Hi folks,
> >
> > I am planning to do some performance research on iSCSI disks mounted
> > on RHEL 3.0 Update5.  
> >
> > I am planning to do this test on raw disks i.e. linux partitions.
> >
> > I am using iometer 2004.07.30 version. When I attach dynamo to the
> > iometer it returns the following message,
> >
> > ...Worker 5 running Access Spec: 32K; 75% Read; 0% random
> > Access specifications for test defined.
> > Grunt: Growing grunt data buffer from 32768 to 32768
> > Worker 0 setting targets...
> > Fail to get size for /dev/sdb1
> > Fail to get size for /dev/sdb1
> > Fail to get size for /dev/sdb1
> > Worker 1 setting targets...
> > Fail to get size for /dev/sdc1
> > Fail to get size for /dev/sdc1
> > Fail to get size for /dev/sdc1
> > Worker 2 setting targets...
> > Fail to get size for /dev/sdd1
> > Fail to get size for /dev/sdd1
> > Fail to get size for /dev/sdd1..."
> >
> > But on the iometer I can select them and assign these partitions as
> > raw disks to the workers. When I start the test, it shows as if it
> is
> > running, results display changes indicating very large figures (e.g.
> > 750MB/s/ 22,895 IOPs) etc. which cannot any way be realistic
> figures.
> > After the specified time interval it stops, csv file indicates very
> > large, unrealistic statistics.
> >
> > But during the run no I/O is actually happening as evident by iostat
> > o/p.
> >
> > I am puzzled. Any ideas as to where I have done wrong?
> >
> >
> > Good Luck!
> > Prasad
> >
>
>



-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server.
Download it for free - -and be entered to win a 42" plasma tv or your very
own Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user
Loading...