Interpreting Iometer Output

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

Interpreting Iometer Output

Emre BALCI
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.

Best Regards

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Wesley_Vaske

Hey Emre,

 

How long did the test run?

 

Wes Vaske

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output

 

    Hi All

I am making simple test on my desktop computer that has only one sata disk and 8GB memory

So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.

For only one sata disk this value is imposibble

where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.

 

Best Regards


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Emre BALCI
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.

Best Regards



From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

ScreenClip.png (53K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Interpreting Iometer Output

Jose Medeiros-2
In reply to this post by Wesley_Vaske
I had similar issue's with Iometer, a few years back, and just assumend that I did not know how to properly read the results. You would think that the tool would offer an option to switch between bits and bytes, and may be such an option is there and I just need a tutorial?

Regrettably the hard drive and storage manaufactures do not have a standard  that's published using Iometer or another industry standarize tool to measure I/O capability, that we can measure against once their products such as Storage Area Networks are implemented.

Sincerely,

Jose F. Medeiros
http://www.linkedin.com/in/josemedeiros




From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 6:45 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Emre BALCI
Hii I thought as you said "am I wrong when reading result are there some mistakens about bits or bytes" but iops just a number cant be this value for only one sata drive.



From: Jose F. Medeiros <[hidden email]>
To: "[hidden email]" <[hidden email]>
Cc: "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>; Jose Medeiros <[hidden email]>
Sent: Thursday, May 3, 2012 5:41 PM
Subject: Re: [Iometer-user] Interpreting Iometer Output

I had similar issue's with Iometer, a few years back, and just assumend that I did not know how to properly read the results. You would think that the tool would offer an option to switch between bits and bytes, and may be such an option is there and I just need a tutorial?

Regrettably the hard drive and storage manaufactures do not have a standard  that's published using Iometer or another industry standarize tool to measure I/O capability, that we can measure against once their products such as Storage Area Networks are implemented.

Sincerely,

Jose F. Medeiros
http://www.linkedin.com/in/josemedeiros




From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 6:45 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user





------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Neto, Antonio Jose Rodrigues
In reply to this post by Emre BALCI

Hi Emre,

 

This is neto from Brazil

 

How are you?

 

You have been raised a very good question.

 

Probably you are facing the issue where you are reading from filesystem cache or memory.

 

Please do this:

 

If you CAN, use physical drive (not filesystem)

Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)

 

The problem that you are facing that your working set is small.

 

 

If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.

 

Please let me know if I can help you with something

 

All the best

 

neto

NetApp – I love this company!

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Wesley

I did not care but at least taken 3-5 minute.

Also you can see test output in the attachment.

 

Best Regards

 

<img id="_x0000_i1025" src="file:///C:\Users\emreb\AppData\Local\Temp\enhtmlclip\ScreenClip(97).png">

 


From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Hey Emre,

 

How long did the test run?

 

Wes Vaske

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output

 

    Hi All

I am making simple test on my desktop computer that has only one sata disk and 8GB memory

So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.

For only one sata disk this value is imposibble

where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.

 

Best Regards

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Jose Medeiros-2
Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.


http://support.microsoft.com/kb/253251


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Jose Medeiros-2
Hi Emre,

Hmm.. that's interesting. Did you also disable write caching on the hard drive using device manager ?

Honestly, one way I was able to accurately verify write performance on a Compaq EMA SAN when I worked at Intel was to create a 4GB file, and move it from one SAN LUN to another SAN LUN and timed how long it took. 

Jose F. Medeiros
http://www.linkedin.com/in/josemedeiros



From: Emre BALCI <[hidden email]>
To: Jose F. Medeiros <[hidden email]>; "Neto, Antonio Jose Rodrigues" <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 11:30 PM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

Jose I tried both enabled and disabled but nothing has changed

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

Wesley On another machine(xp sp2) results still same 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)



From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.


http://support.microsoft.com/kb/253251


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user







------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Neto, Antonio Jose Rodrigues
In reply to this post by Jose Medeiros-2
Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

Jose I tried both enabled and disabled but nothing has changed

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

Wesley On another machine(xp sp2) results still same 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)



From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.


http://support.microsoft.com/kb/253251


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user





------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Emre BALCI
for now is not possible
Did you the steps that I have follwed is there any wrong ??


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

Jose I tried both enabled and disabled but nothing has changed

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

Wesley On another machine(xp sp2) results still same 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)



From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.


http://support.microsoft.com/kb/253251


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user







------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Neto, Antonio Jose Rodrigues
Try to use Rc.1 (a lot of improvements)

You are reading from cache. 100% sure

Always I use raw device to have 100% of sure that I don't have any kind of FS cache 

Try raw disk

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

for now is not possible
Did you the steps that I have follwed is there any wrong ??


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

Jose I tried both enabled and disabled but nothing has changed

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

Wesley On another machine(xp sp2) results still same 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)



From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.


http://support.microsoft.com/kb/253251


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user







------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Emre BALCI
Hii Neto
I am using rc1
I have just tested on raw volumes results still same
Steps are in the attachments


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

Try to use Rc.1 (a lot of improvements)

You are reading from cache. 100% sure

Always I use raw device to have 100% of sure that I don't have any kind of FS cache 

Try raw disk

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

for now is not possible
Did you the steps that I have follwed is there any wrong ??


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

Jose I tried both enabled and disabled but nothing has changed

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

Wesley On another machine(xp sp2) results still same 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)



From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.


http://support.microsoft.com/kb/253251


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user









------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

raw3.png (72K) Download Attachment
raw2.png (62K) Download Attachment
raw1.png (21K) Download Attachment
raw4.png (62K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Interpreting Iometer Output

Neto, Antonio Jose Rodrigues

Please align sectors in 4096 and try again.

 

Is it possible to me to connect remotely on your machine?

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 7:38 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Neto

I am using rc1

I have just tested on raw volumes results still same

Steps are in the attachments

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Try to use Rc.1 (a lot of improvements)

 

You are reading from cache. 100% sure

 

Always I use raw device to have 100% of sure that I don't have any kind of FS cache 

 

Try raw disk


From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

for now is not possible

Did you the steps that I have follwed is there any wrong ??

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Is it possible to try raw device? (disk without volume)


From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

 

Jose I tried both enabled and disabled but nothing has changed

 

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

 

Wesley On another machine(xp sp2) results still same 

 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)

 

 


From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.



diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hi Emre,

 

This is neto from Brazil

 

How are you?

 

You have been raised a very good question.

 

Probably you are facing the issue where you are reading from filesystem cache or memory.

 

Please do this:

 

If you CAN, use physical drive (not filesystem)

Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)

 

The problem that you are facing that your working set is small.

 

 

If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.

 

Please let me know if I can help you with something

 

All the best

 

neto

NetApp – I love this company!

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Wesley

I did not care but at least taken 3-5 minute.

Also you can see test output in the attachment.

 

Best Regards

 

 


From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Hey Emre,

 

How long did the test run?

 

Wes Vaske

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output

 

    Hi All

I am making simple test on my desktop computer that has only one sata disk and 8GB memory

So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.

For only one sata disk this value is imposibble

where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.

 

Best Regards

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

 

 

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Emre BALCI
Results are in attachments



From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 3:54 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

Please align sectors in 4096 and try again.
 
Is it possible to me to connect remotely on your machine?
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 7:38 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Neto
I am using rc1
I have just tested on raw volumes results still same
Steps are in the attachments
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Try to use Rc.1 (a lot of improvements)
 
You are reading from cache. 100% sure
 
Always I use raw device to have 100% of sure that I don't have any kind of FS cache 
 
Try raw disk

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
for now is not possible
Did you the steps that I have follwed is there any wrong ??
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
Hii
 
Jose I tried both enabled and disabled but nothing has changed
 
Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same
 
Wesley On another machine(xp sp2) results still same 
 
I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)
 
 

From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.

http://support.microsoft.com/kb/253251
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

 
 
 



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

test1.png (104K) Download Attachment
test2.png (59K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Interpreting Iometer Output

Neto, Antonio Jose Rodrigues

You put 4MB and 96 bytes

 

Put like this:

 

Megabytes 0

Kilobytes 4

Bytes 0

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 9:09 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Results are in attachments

 

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 3:54 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Please align sectors in 4096 and try again.

 

Is it possible to me to connect remotely on your machine?

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 7:38 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Neto

I am using rc1

I have just tested on raw volumes results still same

Steps are in the attachments

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Try to use Rc.1 (a lot of improvements)

 

You are reading from cache. 100% sure

 

Always I use raw device to have 100% of sure that I don't have any kind of FS cache 

 

Try raw disk


From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

for now is not possible

Did you the steps that I have follwed is there any wrong ??

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Is it possible to try raw device? (disk without volume)


From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

 

Jose I tried both enabled and disabled but nothing has changed

 

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

 

Wesley On another machine(xp sp2) results still same 

 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)

 

 


From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.



diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hi Emre,

 

This is neto from Brazil

 

How are you?

 

You have been raised a very good question.

 

Probably you are facing the issue where you are reading from filesystem cache or memory.

 

Please do this:

 

If you CAN, use physical drive (not filesystem)

Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)

 

The problem that you are facing that your working set is small.

 

 

If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.

 

Please let me know if I can help you with something

 

All the best

 

neto

NetApp – I love this company!

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Wesley

I did not care but at least taken 3-5 minute.

Also you can see test output in the attachment.

 

Best Regards

 

 


From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Hey Emre,

 

How long did the test run?

 

Wes Vaske

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output

 

    Hi All

I am making simple test on my desktop computer that has only one sata disk and 8GB memory

So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.

For only one sata disk this value is imposibble

where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.

 

Best Regards

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

 

 

 

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Emre BALCI


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:11 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

You put 4MB and 96 bytes
 
Put like this:
 
Megabytes 0
Kilobytes 4
Bytes 0
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 9:09 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Results are in attachments
 
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 3:54 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Please align sectors in 4096 and try again.
 
Is it possible to me to connect remotely on your machine?
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 7:38 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Neto
I am using rc1
I have just tested on raw volumes results still same
Steps are in the attachments
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Try to use Rc.1 (a lot of improvements)
 
You are reading from cache. 100% sure
 
Always I use raw device to have 100% of sure that I don't have any kind of FS cache 
 
Try raw disk

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
for now is not possible
Did you the steps that I have follwed is there any wrong ??
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
Hii
 
Jose I tried both enabled and disabled but nothing has changed
 
Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same
 
Wesley On another machine(xp sp2) results still same 
 
I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)
 
 

From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user
 
 
 
 



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

test11.png (107K) Download Attachment
test22.png (58K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Interpreting Iometer Output

Neto, Antonio Jose Rodrigues
I believe that I understood now.

You are doing sequential IO. Sequential IO is easy for any disk.

Change to random and you will see :-)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 9:27 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output



From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:11 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

You put 4MB and 96 bytes
 
Put like this:
 
Megabytes 0
Kilobytes 4
Bytes 0
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 9:09 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Results are in attachments
 
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 3:54 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Please align sectors in 4096 and try again.
 
Is it possible to me to connect remotely on your machine?
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 7:38 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Neto
I am using rc1
I have just tested on raw volumes results still same
Steps are in the attachments
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Try to use Rc.1 (a lot of improvements)
 
You are reading from cache. 100% sure
 
Always I use raw device to have 100% of sure that I don't have any kind of FS cache 
 
Try raw disk

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
for now is not possible
Did you the steps that I have follwed is there any wrong ??
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Is it possible to try raw device? (disk without volume)

From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
Hii
 
Jose I tried both enabled and disabled but nothing has changed
 
Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same
 
Wesley On another machine(xp sp2) results still same 
 
I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)
 
 

From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.


diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.
 

From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hi Emre,
 
This is neto from Brazil
 
How are you?
 
You have been raised a very good question.
 
Probably you are facing the issue where you are reading from filesystem cache or memory.
 
Please do this:
 
If you CAN, use physical drive (not filesystem)
Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)
 
The problem that you are facing that your working set is small.
 
 
If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.
 
Please let me know if I can help you with something
 
All the best
 
neto
NetApp – I love this company!
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output
 
Hii Wesley
I did not care but at least taken 3-5 minute.
Also you can see test output in the attachment.
 
Best Regards
 
 

From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output
 
Hey Emre,
 
How long did the test run?
 
Wes Vaske
 
From: Emre BALCI [mailto:[hidden email]]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output
 
    Hi All
I am making simple test on my desktop computer that has only one sata disk and 8GB memory
So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.
For only one sata disk this value is imposibble
where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.
 
Best Regards
 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user
 
 
 
 



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
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: Interpreting Iometer Output

Wesley_Vaske
In reply to this post by Emre BALCI

Hey Emre,

 

That shows 0% random. When you’re doing sequential IO the numbers you’re getting are what you would expect from your drive.

 

Wes Vaske

 

From: Emre BALCI [mailto:[hidden email]]
Sent: Friday, May 04, 2012 6:38 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; Vaske, Wes; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Neto

I am using rc1

I have just tested on raw volumes results still same

Steps are in the attachments

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:08 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Try to use Rc.1 (a lot of improvements)

 

You are reading from cache. 100% sure

 

Always I use raw device to have 100% of sure that I don't have any kind of FS cache 

 

Try raw disk


From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 7:06 AM
To: Neto, Antonio Jose Rodrigues; Jose F. Medeiros; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

for now is not possible

Did you the steps that I have follwed is there any wrong ??

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; Jose F. Medeiros <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 2:03 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Is it possible to try raw device? (disk without volume)


From: Emre BALCI [[hidden email]]
Sent: Friday, May 04, 2012 2:30 AM
To: Jose F. Medeiros; Neto, Antonio Jose Rodrigues; [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

Hii

 

Jose I tried both enabled and disabled but nothing has changed

 

Neto I tried as you said iometer consume all free space on disk driver that is 26 GB space but results still same

 

Wesley On another machine(xp sp2) results still same 

 

I cant belive that where I am doing wrong ??The steps those I have followed in the attachments. I am just selecting one worker process,maximum disk size 30000000 (It creates 15GB iobw.tst file)

 

 


From: Jose F. Medeiros <[hidden email]>
To: "Neto, Antonio Jose Rodrigues" <[hidden email]>; Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Friday, May 4, 2012 4:35 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Have you tried enabling / disabling counters using Diskperf? I am not sure if matters using Iometer, but it's recommended by Microsoft when using Perfmon.



diskperf

Both Logical and Physical Disk Performance counters on this system
        are automatically enabled on demand.
For legacy applications using IOCTL_DISK_PERFORMANCE to retrieve raw counters,
you can use -Y or -N to forcibly enable or disable.

 


From: "Neto, Antonio Jose Rodrigues" <[hidden email]>
To: Emre BALCI <[hidden email]>; "[hidden email]" <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Thursday, May 3, 2012 10:52 AM
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hi Emre,

 

This is neto from Brazil

 

How are you?

 

You have been raised a very good question.

 

Probably you are facing the issue where you are reading from filesystem cache or memory.

 

Please do this:

 

If you CAN, use physical drive (not filesystem)

Use a big working set don’t specify any values (if you use an small working your numbers are coming from cache)

 

The problem that you are facing that your working set is small.

 

 

If you have iobs.tst, please remove it, create a file bigger than 8GB and run the tests again.

 

Please let me know if I can help you with something

 

All the best

 

neto

NetApp – I love this company!

 

From: Emre BALCI [hidden email]
Sent: Thursday, May 03, 2012 10:25 AM
To: [hidden email]; [hidden email]
Subject: Re: [Iometer-user] Interpreting Iometer Output

 

Hii Wesley

I did not care but at least taken 3-5 minute.

Also you can see test output in the attachment.

 

Best Regards

 

 


From: "[hidden email]" <[hidden email]>
To: [hidden email]; [hidden email]
Sent: Thursday, May 3, 2012 4:45 PM
Subject: RE: [Iometer-user] Interpreting Iometer Output

 

Hey Emre,

 

How long did the test run?

 

Wes Vaske

 

From: Emre BALCI [hidden email]
Sent: Thursday, May 03, 2012 3:09 AM
To: [hidden email]
Subject: [Iometer-user] Interpreting Iometer Output

 

    Hi All

I am making simple test on my desktop computer that has only one sata disk and 8GB memory

So I chooese max disk size 30048000 and 4k Random %100 read test and iometer shows me 6900 for Total io and 0.1437 for avg response time.

For only one sata disk this value is imposibble

where am I wrong ? If I choose default test then values seems ok for only one sata disk but I need make sure for 4k Random %100 read test is correct before make test on SAN.

 

Best Regards

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

 

 

 


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

raw3.png (72K) Download Attachment
Loading...