IOmeter contact id regarding issues

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

IOmeter contact id regarding issues

Elavazhagan.K

Dear Sir,

            We are using Io meter tool for measuring HDD disk & LAN performance. We have some doubts and query regarding this tool. Can we write the query mail to this mail id or not? If not is there any contact mail id to clarify our issues regarding IOmeter?

 

With Regards,

Elavazhagan.K


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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 contact id regarding issues

Ming Zhang
On Thu, 2007-02-08 at 16:40 +0530, Elavazhagan.K wrote:
> Dear Sir,
>
>             We are using Io meter tool for measuring HDD disk & LAN
> performance. We have some doubts and query regarding this tool. Can we
> write the query mail to this mail id or not? If not is there any
> contact mail id to clarify our issues regarding IOmeter?
>

this is iometer user list. the right place for you.

Ming

>  
>
> With Regards,
>
> Elavazhagan.K
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier.
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________ Iometer-user mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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

queris in IOmeter

Elavazhagan.K

Dear Ming,

               

                We have some following queries and doubts in IOmeter. Please clarify us.

 

  1. We are using IOmeter version 2004.07.30. What is latest version of IOmeter & where can we get it? Please provide the URL to download.
  2. One motherboard has two LAN controllers 1) L1- Intel 2) L2-Marvel. We connected two same mother board point to point and did the benchmarking. Please see the results below. From this how to conclude which controller is performing good?

connected between Machine 1-machine2

L1 - Intel LAN

 

L2 - Marvel LAN

 

L1-L1

210 MBs

L2-L1

206 MBs

L2-L2

222 MBs

L1-L2

223 MBs







 

 

  1. We run IOmeter for LAN performance in one motherboard. After the results have taken we compared the scores, in that we have some confusion.

a.        Is LAN performance dependence upon the destination LAN controller? If yes means how to compare the onboard LAN performances of two motherboards with same destination LAN controller?

b.       In the below results CPU utilization are high in all the three. Why we go for LAN card?

c.        We used here all 1Gbps LAN controller i.e.  It can able to transfer 125 MBs per second. How the score shows 213 MBs?

 

1

Asus + Onboard LAN (10/100/1000)Mbps

 (10MB request size, 100% read, 100 sequent ional)

Total I/O 's per second

17.91

Total MB's per second

179.12

Average I/O response time

111.65

Maximum I/O response time(ms)

173.82

% CPU Utilization

70.6

2

Asus + PCI 32 bit Dragon LAN card

(10MB request size, 100% read, 100 sequent ional)

Total I/O 's per second

21.31

Total MB's per second

213.15

Average I/O response time

93.82

Maximum I/O response time(ms)

172.03

% CPU Utilization

62.12

3

Asus + ASUS PCI-Express Gigabyte LAN card PEB-G21

(10MB request size, 100% read, 100 sequent ional)

Total I/O 's per second

21.31

Total MB's per second

213.09

Average I/O response time

93.849

Maximum I/O response time(ms)

168.86

% CPU Utilization

61.86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Please clarify us.

 

With regards,

Elavazhagan.K

 

 

 

 

 

















 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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: queris in IOmeter

Ming Zhang
On Fri, 2007-02-09 at 15:45 +0530, Elavazhagan.K wrote:

> Dear Ming,
>
>                
>
>                 We have some following queries and doubts in IOmeter.
> Please clarify us.
>
>  
>
>      1. We are using IOmeter version 2004.07.30. What is latest
>         version of IOmeter & where can we get it? Please provide the
>         URL to download.

search in sf.net

followed message are severely skewed. pls repost. do not use html email,
please.

 

>      1. One motherboard has two LAN controllers 1) L1- Intel 2)
>         L2-Marvel. We connected two same mother board point to point
>         and did the benchmarking. Please see the results below. From
>         this how to conclude which controller is performing good?
>    connected between
>   Machine 1-machine2
>
>
>     L1 - Intel LAN
>
>
>            
>
>
>     L2 - Marvel LAN
>
>
>            
>
>
>          L1-L1
>
>
>         210 MBs
>
>
>          L2-L1
>
>
>         206 MBs
>
>
>          L2-L2
>
>
>         222 MBs
>
>
>          L1-L2
>
>
>         223 MBs
>
>
>
>
>
>
>
>
>
>  
>
>  
>
>      1. We run IOmeter for LAN performance in one motherboard. After
>         the results have taken we compared the scores, in that we have
>         some confusion.
>
> a.       Is LAN performance dependence upon the destination LAN
> controller? If yes means how to compare the onboard LAN performances
> of two motherboards with same destination LAN controller?
>
> b.       In the below results CPU utilization are high in all the
> three. Why we go for LAN card?
>
> c.       We used here all 1Gbps LAN controller i.e.  It can able to
> transfer 125 MBs per second. How the score shows 213 MBs?
>
>  
>
> 1
>
>
>  Asus + Onboard
>        LAN
> (10/100/1000)Mbps
>
>   (10MB request
> size, 100% read,
>    100 sequent
>      ional)
>
>
> Total I/O 's per
> second
>
>
>       17.91
>
>
> Total MB's per
> second
>
>
>      179.12
>
>
> Average I/O
> response time
>
>
>      111.65
>
>
> Maximum I/O
> response time(ms)
>
>
>      173.82
>
>
> % CPU Utilization
>
>
>       70.6
>
>
>         2
>
>
> Asus + PCI 32 bit
> Dragon LAN card
>
>   (10MB request
> size, 100% read,
>    100 sequent
>      ional)
>
>
> Total I/O 's per
> second
>
>
>       21.31
>
>
> Total MB's per
> second
>
>
>      213.15
>
>
> Average I/O
> response time
>
>
>       93.82
>
>
> Maximum I/O
> response time(ms)
>
>
>      172.03
>
>
> % CPU Utilization
>
>
>       62.12
>
>
>         3
>
>
>    Asus + ASUS
>    PCI-Express
> Gigabyte LAN card
>     PEB-G21
>
>   (10MB request
> size, 100% read,
>    100 sequent
>      ional)
>
>
> Total I/O 's per
> second
>
>
>       21.31
>
>
> Total MB's per
> second
>
>
>      213.09
>
>
> Average I/O
> response time
>
>
>      93.849
>
>
> Maximum I/O
> response time(ms)
>
>
>      168.86
>
>
> % CPU Utilization
>
>
>       61.86
>
>
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
> Please clarify us.
>
>  
>
> With regards,
>
> Elavazhagan.K
>
>  
>
>  
>
>  
>
>  
>
>  
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>  
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier.
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________ Iometer-user mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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: queris in IOmeter

Vedran Degoricija
In reply to this post by Elavazhagan.K
Please keep in mind that iometer reports the sum of both endpoint results. The dynamo on each side is seeing ~110MB/s, so your total is ~220MB/s.
 
Regards,
Ved

----- Original Message ----
From: Elavazhagan.K <[hidden email]>
To: [hidden email]
Cc: [hidden email]
Sent: Friday, February 9, 2007 2:15:35 AM
Subject: [Iometer-user] queris in IOmeter

Dear Ming,

               

                We have some following queries and doubts in IOmeter. Please clarify us.

 

  1. We are using IOmeter version 2004.07.30. What is latest version of IOmeter & where can we get it? Please provide the URL to download.
  2. One motherboard has two LAN controllers 1) L1- Intel 2) L2-Marvel. We connected two same mother board point to point and did the benchmarking. Please see the results below. From this how to conclude which controller is performing good?

connected between Machine 1-machine2

L1 - Intel LAN

 

L2 - Marvel LAN

 

L1-L1

210 MBs

L2-L1

206 MBs

L2-L2

222 MBs

L1-L2

223 MBs







 

 

  1. We run IOmeter for LAN performance in one motherboard. After the results have taken we compared the scores, in that we have some confusion.

a.        Is LAN performance dependence upon the destination LAN controller? If yes means how to compare the onboard LAN performances of two motherboards with same destination LAN controller?

b.       In the below results CPU utilization are high in all the three. Why we go for LAN card?

c.        We used here all 1Gbps LAN controller i.e.  It can able to transfer 125 MBs per second. How the score shows 213 MBs?

 

1

Asus + Onboard LAN (10/100/1000)Mbps

 (10MB request size, 100% read, 100 sequent ional)

Total I/O 's per second

17.91

Total MB's per second

179.12

Average I/O response time

111.65

Maximum I/O response time(ms)

173.82

% CPU Utilization

70.6

2

Asus + PCI 32 bit Dragon LAN card

(10MB request size, 100% read, 100 sequent ional)

Total I/O 's per second

21.31

Total MB's per second

213.15

Average I/O response time

93.82

Maximum I/O response time(ms)

172.03

% CPU Utilization

62.12

3

Asus + ASUS PCI-Express Gigabyte LAN card PEB-G21

(10MB request size, 100% read, 100 sequent ional)

Total I/O 's per second

21.31

Total MB's per second

213.09

Average I/O response time

93.849

Maximum I/O response time(ms)

168.86

% CPU Utilization

61.86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Please clarify us.

 

With regards,

Elavazhagan.K

 

 

 

 

 

















 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user



Check out the all-new Yahoo! Mail beta - Fire up a more powerful email and get things done faster.
-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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

queris in IOmeter-LAN performance

Elavazhagan.K
In reply to this post by Ming Zhang
Dear Ming,

IOmeter Queries:

1. One motherboard has two LAN controllers 1) L1-Intel 2) L2-Marvel.We
connected the two same mother board point to point and did the benchmarking.
Please see the results below. From this how to conclude which controller is
performing good?

======================================
Connected between           scores
======================================
1) L1 to L1    210 MBs
2) L1 to L2                 223 MBs
3) L2 to L1                 206 MBs
4) L2 to L2                 222 MBs
=====================================

2. We run IOmeter for LAN performance in one motherboard. After the results
have taken we compared the scores, in that we have some confusion.
        a) Is LAN performance dependence upon the destination LAN
controller? If yes means how to compare the onboard LAN performances of
different motherboards with same destination LAN controller?
        b) If we compared the results of Onboard LAN, PCI LAN card & PCI
Express LAN Card. All the results are giving CPU utilization is high. Please
say some comments on this.
        c) We find no difference in benchmarking results if we compare the
results of PCI card Vs PCI express card. How PCI card can perform equal to
PCI Express card? Is it a tool problem?

Please resolve these queries.

With Regards,
Elavazhagan.K

 
       


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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: queris in IOmeter-LAN performance

Elavazhagan.K
Dear Ming,
Please revert on this.

With Regards,
Elavazhagan.K

-----Original Message-----
From: Elavazhagan.K [mailto:[hidden email]]
Sent: Saturday, February 10, 2007 9:28 AM
To: '[hidden email]'
Cc: '[hidden email]'
Subject: queris in IOmeter-LAN performance

Dear Ming,

IOmeter Queries:

1. One motherboard has two LAN controllers 1) L1-Intel 2) L2-Marvel.We
connected the two same mother board point to point and did the benchmarking.
Please see the results below. From this how to conclude which controller is
performing good?

======================================
Connected between           scores
======================================
1) L1 to L1    210 MBs
2) L1 to L2                 223 MBs
3) L2 to L1                 206 MBs
4) L2 to L2                 222 MBs
=====================================

2. We run IOmeter for LAN performance in one motherboard. After the results
have taken we compared the scores, in that we have some confusion.
        a) Is LAN performance dependence upon the destination LAN
controller? If yes means how to compare the onboard LAN performances of
different motherboards with same destination LAN controller?
        b) If we compared the results of Onboard LAN, PCI LAN card & PCI
Express LAN Card. All the results are giving CPU utilization is high. Please
say some comments on this.
        c) We find no difference in benchmarking results if we compare the
results of PCI card Vs PCI express card. How PCI card can perform equal to
PCI Express card? Is it a tool problem?

Please resolve these queries.

With Regards,
Elavazhagan.K

 
       


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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: queris in IOmeter-LAN performance

Ming Zhang
On Tue, 2007-02-13 at 12:44 +0530, Elavazhagan.K wrote:
> Dear Ming,
> Please revert on this.

revert on this? i am confused.


>
> With Regards,
> Elavazhagan.K
>
> -----Original Message-----
> From: Elavazhagan.K [mailto:[hidden email]]
> Sent: Saturday, February 10, 2007 9:28 AM
> To: '[hidden email]'
> Cc: '[hidden email]'
> Subject: queris in IOmeter-LAN performance
>
> Dear Ming,
>
> IOmeter Queries:
>
> 1. One motherboard has two LAN controllers 1) L1-Intel 2) L2-Marvel.We
> connected the two same mother board point to point and did the benchmarking.
> Please see the results below. From this how to conclude which controller is
> performing good?

what is your definition on "perform good"?

>
> ======================================
> Connected between           scores
> ======================================
> 1) L1 to L1    210 MBs
> 2) L1 to L2                 223 MBs
> 3) L2 to L1                 206 MBs
> 4) L2 to L2                 222 MBs
> =====================================
>
> 2. We run IOmeter for LAN performance in one motherboard. After the results
> have taken we compared the scores, in that we have some confusion.
> a) Is LAN performance dependence upon the destination LAN
> controller? If yes means how to compare the onboard LAN performances of

yes. it will depends on interconnection, NIC, host cpu, memory
bandwidth,...

> different motherboards with same destination LAN controller?
> b) If we compared the results of Onboard LAN, PCI LAN card & PCI
> Express LAN Card. All the results are giving CPU utilization is high. Please
> say some comments on this.

check u motherboard manual, what is the difference between a onboard LAN
and plugged external NIC?

> c) We find no difference in benchmarking results if we compare the
> results of PCI card Vs PCI express card. How PCI card can perform equal to

possible.

> PCI Express card? Is it a tool problem?

why u think PCIe __MUST__ perform better than PCI?

please check more information on what is the difference between PCIe and
PCI then you will answer your own question.

>
> Please resolve these queries.
>
> With Regards,
> Elavazhagan.K
>
>  
>
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier.
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
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: queris in IOmeter-LAN performance

Elavazhagan.K
Dear Ming,

Query:
        One motherboard has two LAN controllers 1) L1-Intel 2) L2-Marvel.We
        Connected the two same mother board LAN Port point to point and did
the benchmarking. Please see the results below.

                ======================================
  Connected between           scores
          ======================================
          1) L1 to L1    210 MBs
  2) L1 to L2                 223 MBs
  3) L2 to L1                 206 MBs
                4) L2 to L2                 222 MBs
  =====================================
>From this how to conclude which controller is having high performance?

Thanks & Regards,
Elavazhagan.K


-----Original Message-----
From: Ming Zhang [mailto:[hidden email]]
Sent: Tuesday, February 13, 2007 7:21 PM
To: Elavazhagan.K
Cc: [hidden email]
Subject: Re: [Iometer-user] queris in IOmeter-LAN performance

On Tue, 2007-02-13 at 12:44 +0530, Elavazhagan.K wrote:
> Dear Ming,
> Please revert on this.

revert on this? i am confused.


>
> With Regards,
> Elavazhagan.K
>
> -----Original Message-----
> From: Elavazhagan.K [mailto:[hidden email]]
> Sent: Saturday, February 10, 2007 9:28 AM
> To: '[hidden email]'
> Cc: '[hidden email]'
> Subject: queris in IOmeter-LAN performance
>
> Dear Ming,
>
> IOmeter Queries:
>
> 1. One motherboard has two LAN controllers 1) L1-Intel 2) L2-Marvel.We
> connected the two same mother board point to point and did the
benchmarking.
> Please see the results below. From this how to conclude which controller
is
> performing good?

what is your definition on "perform good"?

>
> ======================================
> Connected between           scores
> ======================================
> 1) L1 to L1    210 MBs
> 2) L1 to L2                 223 MBs
> 3) L2 to L1                 206 MBs
> 4) L2 to L2                 222 MBs
> =====================================
>
> 2. We run IOmeter for LAN performance in one motherboard. After the
results
> have taken we compared the scores, in that we have some confusion.
> a) Is LAN performance dependence upon the destination LAN
> controller? If yes means how to compare the onboard LAN performances of

yes. it will depends on interconnection, NIC, host cpu, memory
bandwidth,...

> different motherboards with same destination LAN controller?
> b) If we compared the results of Onboard LAN, PCI LAN card & PCI
> Express LAN Card. All the results are giving CPU utilization is high.
Please
> say some comments on this.

check u motherboard manual, what is the difference between a onboard LAN
and plugged external NIC?

> c) We find no difference in benchmarking results if we compare the
> results of PCI card Vs PCI express card. How PCI card can perform equal to

possible.

> PCI Express card? Is it a tool problem?

why u think PCIe __MUST__ perform better than PCI?

please check more information on what is the difference between PCIe and
PCI then you will answer your own question.

>
> Please resolve these queries.
>
> With Regards,
> Elavazhagan.K
>
>  
>
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job
easier.
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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

Query regarding sata HDD

Elavazhagan.K
In reply to this post by Ming Zhang
Dear Ming,

We did Benchmarking for Sata Hard disk drive using IO-Meter Tool.

Query:
======
Western Digital 250GB 7200rpm sata/300 8MB buffer (WD2500JS)
============================================================================
IO Meter Tool
64KB Transfer request size, 100% read & Sequential
============================================================================
Total I/Os per Second 385.76
Total MB/s per Second 24.11
Average I/O response time (ms) 5.18
Maximum I/O response time (ms) 17.60
% CPU utilization 0.81
============================================================================

Normally sata HDD has the capable of transferring 60MB/s. In the above
results, Total MB/s shows only 24.11. If we run with some other tool like
HDtach & HD Tune tool, it is giving 54 MB/s.

Can we know why this tool showing low values (that to only for few hard disk
drives)?
Whether is this tool problem?
How can I believe that this tool showing correct value?
Can I get the procedure to benchmark how you do in your LAB?

Please help us to solve these queries.


With Regards,
Elavazhagan.K



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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

percentage counter error in IOmeter

Elavazhagan.K
In reply to this post by Ming Zhang
Dear Ming,

When doing Benchmarking for HDD, some times the below error is coming.

*** Performance counter percentage is less than zero:-0.00893216
*** Performance counter percentage is less than zero:-0.00893216
*** Performance counter percentage is less than zero:-0.00893216
*** Performance counter percentage is less than zero:-0.00893216


What is the meaning of this error? Can you say in brief?
What actually happens in background at time of running IOmeter Tool?
How can I prove that this tool giving the correct value to my costumer?

If this error came, can we conclude the HDD is not performing well?

With Regards,
Elavazhagan.K


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user

query in iometer.JPG (140K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: percentage counter error in IOmeter

Daniel Scheibli-2

Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:

> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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: Query regarding sata HDD

Elavazhagan.K
In reply to this post by Elavazhagan.K
Dear Sir,

        Please answer for the below queries also.

With Regards,
Elavazhagan.K

-----Original Message-----
From: elavazhagan.k [mailto:[hidden email]]
Sent: Saturday, March 24, 2007 2:45 PM
To: '[hidden email]'
Cc: '[hidden email]'
Subject: Query regarding sata HDD

Dear Ming,

We did Benchmarking for Sata Hard disk drive using IO-Meter Tool.

Query:
======
Western Digital 250GB 7200rpm sata/300 8MB buffer (WD2500JS)
============================================================================
IO Meter Tool
64KB Transfer request size, 100% read & Sequential
============================================================================
Total I/Os per Second 385.76
Total MB/s per Second 24.11
Average I/O response time (ms) 5.18
Maximum I/O response time (ms) 17.60
% CPU utilization 0.81
============================================================================

Normally sata HDD has the capable of transferring 60MB/s. In the above
results, Total MB/s shows only 24.11. If we run with some other tool like
HDtach & HD Tune tool, it is giving 54 MB/s.

Can we know why this tool showing low values (that to only for few hard disk
drives)?
Whether is this tool problem?
How can I believe that this tool showing correct value?
Can I get the procedure to benchmark how you do in your LAB?

Please help us to solve these queries.


With Regards,
Elavazhagan.K



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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: percentage counter error in IOmeter

Elavazhagan.K
In reply to this post by Daniel Scheibli-2
Dear Daniel,

Thanks for your kind reply. As of now we removed the testing bed.
We could not able to give the information what you want. If this problem
occurs another time let me intimate to you.

Is there any procedure or precautions to avoid this error?

With Regards,
Elavazhagan.K



-----Original Message-----
From: Daniel Scheibli [mailto:[hidden email]]
Sent: Sunday, March 25, 2007 11:46 PM
To: elavazhagan.k
Cc: [hidden email]
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:

> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
your

> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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: percentage counter error in IOmeter

Vedran Degoricija
In reply to this post by Elavazhagan.K
Elavazhagan,

Unless this error happens constantly during the test, it should not really be of much concern. Can you tell me if it happens when you stop the worker and restart it again?

Also, for your performance problem, what are the access specification details you are using?

Regards,
Ved

----- Original Message ----
From: elavazhagan.k <[hidden email]>
To: Daniel Scheibli <[hidden email]>
Cc: [hidden email]
Sent: Sunday, March 25, 2007 8:35:06 PM
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Dear Daniel,

Thanks for your kind reply. As of now we removed the testing bed.
We could not able to give the information what you want. If this problem
occurs another time let me intimate to you.

Is there any procedure or precautions to avoid this error?

With Regards,
Elavazhagan.K



-----Original Message-----
From: Daniel Scheibli [mailto:[hidden email]]
Sent: Sunday, March 25, 2007 11:46 PM
To: elavazhagan.k
Cc: [hidden email]
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:

> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
your

> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user


 
____________________________________________________________________________________
Get your own web address.  
Have a HUGE year through Yahoo! Small Business.
http://smallbusiness.yahoo.com/domains/?p=BESTDEAL

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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: percentage counter error in IOmeter

Elavazhagan.K
Dear Vedran,

We have used 64KB transfer request size, 100 % read & 100 % sequential for
our testing. Will this be also producing this error?


With Regards,
Elavazhaagn.K


-----Original Message-----
From: Vedran Degoricija [mailto:[hidden email]]
Sent: Monday, March 26, 2007 2:10 PM
To: elavazhagan.k; Daniel Scheibli
Cc: [hidden email]
Subject: Re: [Iometer-user] percentage counter error in IOmeter

Elavazhagan,

Unless this error happens constantly during the test, it should not really
be of much concern. Can you tell me if it happens when you stop the worker
and restart it again?

Also, for your performance problem, what are the access specification
details you are using?

Regards,
Ved

----- Original Message ----
From: elavazhagan.k <[hidden email]>
To: Daniel Scheibli <[hidden email]>
Cc: [hidden email]
Sent: Sunday, March 25, 2007 8:35:06 PM
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Dear Daniel,

Thanks for your kind reply. As of now we removed the testing bed.
We could not able to give the information what you want. If this problem
occurs another time let me intimate to you.

Is there any procedure or precautions to avoid this error?

With Regards,
Elavazhagan.K



-----Original Message-----
From: Daniel Scheibli [mailto:[hidden email]]
Sent: Sunday, March 25, 2007 11:46 PM
To: elavazhagan.k
Cc: [hidden email]
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:

> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
your

> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user


 
____________________________________________________________________________
________
Get your own web address.  
Have a HUGE year through Yahoo! Small Business.
http://smallbusiness.yahoo.com/domains/?p=BESTDEAL


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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: percentage counter error in IOmeter

Vedran Degoricija
In reply to this post by Elavazhagan.K
Elavazhaagn,

No, the erors are not related to your poor performance. Like Daniel guessed, it might be related to teh CPUs' timestamp counter. Is this being run on a multi-proc machine or a laptop?

Have you tried increasing the number of outstanding IOs field in iometer gui?

Ved

----- Original Message ----
From: elavazhagan.k <[hidden email]>
To: Vedran Degoricija <[hidden email]>; Daniel Scheibli <[hidden email]>
Cc: [hidden email]
Sent: Monday, March 26, 2007 4:00:02 AM
Subject: RE: [Iometer-user] percentage counter error in IOmeter


Dear Vedran,

We have used 64KB transfer request size, 100 % read & 100 % sequential for
our testing. Will this be also producing this error?


With Regards,
Elavazhaagn.K


-----Original Message-----
From: Vedran Degoricija [mailto:[hidden email]]
Sent: Monday, March 26, 2007 2:10 PM
To: elavazhagan.k; Daniel Scheibli
Cc: [hidden email]
Subject: Re: [Iometer-user] percentage counter error in IOmeter

Elavazhagan,

Unless this error happens constantly during the test, it should not really
be of much concern. Can you tell me if it happens when you stop the worker
and restart it again?

Also, for your performance problem, what are the access specification
details you are using?

Regards,
Ved

----- Original Message ----
From: elavazhagan.k <[hidden email]>
To: Daniel Scheibli <[hidden email]>
Cc: [hidden email]
Sent: Sunday, March 25, 2007 8:35:06 PM
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Dear Daniel,

Thanks for your kind reply. As of now we removed the testing bed.
We could not able to give the information what you want. If this problem
occurs another time let me intimate to you.

Is there any procedure or precautions to avoid this error?

With Regards,
Elavazhagan.K



-----Original Message-----
From: Daniel Scheibli [mailto:[hidden email]]
Sent: Sunday, March 25, 2007 11:46 PM
To: elavazhagan.k
Cc: [hidden email]
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:

> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
your

> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Iometer-user mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Iometer-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/iometer-user



____________________________________________________________________________
________
Get your own web address.  
Have a HUGE year through Yahoo! Small Business.
http://smallbusiness.yahoo.com/domains/?p=BESTDEAL


 
____________________________________________________________________________________
No need to miss a message. Get email on-the-go
with Yahoo! Mail for Mobile. Get started.
http://mobile.yahoo.com/mail 

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
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: percentage counter error in IOmeter

chengguan
Dear sir,

I am facing similar issue elavazhagan faced and I am using Xeon 5160 processor. According to your advice, my processor should be fine. However, I am still seeing the "Performance counter percentage is less than zero" warning!

Thank you!

Regards,
Chengguan.

-----Original Message-----
From: Daniel Scheibli [mailto:daniel@scheibli.com]
Sent: Sunday, March 25, 2007 11:46 PM
To: elavazhagan.k
Cc: iometer-user@lists.sourceforge.net
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:
> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: percentage counter error in IOmeter

AnatolyVilchinsky
In reply to this post by Vedran Degoricija
Does anyone found solution for this issue? I've got the same, so I hope that you'll kindly help me.


Vedran Degoricija wrote
Elavazhaagn,

No, the erors are not related to your poor performance. Like Daniel guessed, it might be related to teh CPUs' timestamp counter. Is this being run on a multi-proc machine or a laptop?

Have you tried increasing the number of outstanding IOs field in iometer gui?

Ved

----- Original Message ----
From: elavazhagan.k <elavazhagan@hcl.in>
To: Vedran Degoricija <vedrand@yahoo.com>; Daniel Scheibli <daniel@scheibli.com>
Cc: iometer-user@lists.sourceforge.net
Sent: Monday, March 26, 2007 4:00:02 AM
Subject: RE: [Iometer-user] percentage counter error in IOmeter


Dear Vedran,

We have used 64KB transfer request size, 100 % read & 100 % sequential for
our testing. Will this be also producing this error?


With Regards,
Elavazhaagn.K


-----Original Message-----
From: Vedran Degoricija [mailto:vedrand@yahoo.com]
Sent: Monday, March 26, 2007 2:10 PM
To: elavazhagan.k; Daniel Scheibli
Cc: iometer-user@lists.sourceforge.net
Subject: Re: [Iometer-user] percentage counter error in IOmeter

Elavazhagan,

Unless this error happens constantly during the test, it should not really
be of much concern. Can you tell me if it happens when you stop the worker
and restart it again?

Also, for your performance problem, what are the access specification
details you are using?

Regards,
Ved

----- Original Message ----
From: elavazhagan.k <elavazhagan@hcl.in>
To: Daniel Scheibli <daniel@scheibli.com>
Cc: iometer-user@lists.sourceforge.net
Sent: Sunday, March 25, 2007 8:35:06 PM
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Dear Daniel,

Thanks for your kind reply. As of now we removed the testing bed.
We could not able to give the information what you want. If this problem
occurs another time let me intimate to you.

Is there any procedure or precautions to avoid this error?

With Regards,
Elavazhagan.K



-----Original Message-----
From: Daniel Scheibli [mailto:daniel@scheibli.com]
Sent: Sunday, March 25, 2007 11:46 PM
To: elavazhagan.k
Cc: iometer-user@lists.sourceforge.net
Subject: Re: [Iometer-user] percentage counter error in IOmeter


Hi Elavazhagan,

can you please provide some more detail on the host - in particular
the CPU?

Sounds like an issue with CPU frequency scaling in combination with
Iometer using the RDTSC instruction for measuring time. For Intel
CPU's you can check chapter 18.10 in the Software Developers Manual
Volume 3B (see [1]) to see if your particular CPU is scaling the
counter based on the CPU frequency (causing problems) or increases
constantly over time (fine).

If you happen to run on a host with the first kind of CPU, then
you should consider doing the testing from another box.

Best Regards
Daniel


[1] http://www.intel.com/products/processor/manuals/



elavazhagan.k wrote:
> Dear Ming,
>
> When doing Benchmarking for HDD, some times the below error is coming.
>
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
> *** Performance counter percentage is less than zero:-0.00893216
>
>
> What is the meaning of this error? Can you say in brief?
> What actually happens in background at time of running IOmeter Tool?
> How can I prove that this tool giving the correct value to my costumer?
>
> If this error came, can we conclude the HDD is not performing well?
>
> With Regards,
> Elavazhagan.K
>
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share
your
> opinions on IT & business topics through brief surveys-and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Iometer-user mailing list
> Iometer-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/iometer-user


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Iometer-user mailing list
Iometer-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/iometer-user



____________________________________________________________________________
________
Get your own web address.  
Have a HUGE year through Yahoo! Small Business.
http://smallbusiness.yahoo.com/domains/?p=BESTDEAL


 
____________________________________________________________________________________
No need to miss a message. Get email on-the-go
with Yahoo! Mail for Mobile. Get started.
http://mobile.yahoo.com/mail 

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Iometer-user mailing list
Iometer-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/iometer-user
Loading...