You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by monpru <pr...@stud.ntnu.no> on 2010/07/08 13:43:29 UTC
negative response times again
Hi,
I got negative latency values in my last load test of a web service. I
searched the web and found that this might have something to do with the jdk
I use? I wonder what the problem is or how to ensure that I get correct
latency in future tests? I am not allowed to use Wireshark or similar to
check transmission times. The load during the tests was quite low (350
requests/minute) and none of the machines involved in the test struggled. I
used 2 thread groups with one Constant Throughput Timer each, all samplers
are SOAP/XML-RPC Requests. I use csv format for logging.
I tried to re-run the test using different jdks and got results that
differed, but had negative and 0 latency in each of the tests.
I used JMeter 2.3.4 r785646
and tried with the following
-Java HotSpot(TM) Server VM (build 1.5.0_16-b02, mixed mode)
-BEA JRockit(R) (build
R26.4.0-63_CR302700-72606-1.5.0_06-20061127-1104-linux-ia32
-Java HotSpot(TM) Server VM (build 1.5.0_06-b05, mixed mode)
on a machine with CentOS release 4.7 and got negative and 0 latency for all.
I tried on a different machine using BEA WebLogic JRockit(TM) 1.4.2_08 JVM
R24.5.0-61 on Red Hat Enterprise Linux ES release 4 (Nahant Update 2): here
I only had 0 load time (the last value) for some requests, but no negative
or 0 latency. But I am unsure if this means the latency is correct this
time?
I might be overlooking something obvious here, so I would be thankful for
hints regarding this problem.
Regards,
Monique
--
View this message in context: http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1044999.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: negative response times again
Posted by monpru <pr...@stud.ntnu.no>.
Hi,
the total amount of negative response times is low, only 3-5 with negative
latency in 40000 responses per test run. The range of the negative values is
from -1163 to -7285 ms. In other tests (same environment) the min. response
times for those operations are about 50 - 200 ms.
Regards,
Monique
--
View this message in context: http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1046322.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: negative response times again
Posted by Deepak Goel <de...@gmail.com>.
Hey Monique
0 Latency is fine on the same subnet considering your accuracy level
is 100 ms. Can you post how much negative latency were you getting?
Thanks
Deepak
On 7/9/10, monpru <pr...@stud.ntnu.no> wrote:
>
> Hi,
> 1. I am not able to get that information since we only get to lend servers
> for testing per project and have little control over the environment. The
> machines are on the same subnet.
> 2. I think it is milliseconds (if I got your question right).
> 3. I ran a shorter test with the result tree listener and it looked ok. The
> responses in the log also look normal (code 200). This is how it looks like
> in the log:
> 2010/07/07 20:10:49.0189,16,Completion,200,Main scenario 1-3,text,true,0,0
> 2010/07/07 20:10:49.0282,64,getStatus,200,Main scenario
> 1-15,text,true,1703,64
> 2010/07/07 20:10:49.0372,-4507,getStatus,200,Main scenario
> 1-12,text,true,1703,-4507
>
> Thank you,
> Monique
> --
> View this message in context:
> http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1045406.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
--
Namaskara~Nalama~Guten Tag
--
Keigu
Deepak
+91-9765089593
deicool@gmail.com
Skype: thumsupdeicool
Google talk: deicool
Blog: http://loveandfearless.wordpress.com
Facebook: http://www.facebook.com/deicool
Check out my Work at:
LinkedIn: http://in.linkedin.com/in/thumsupdeicool
"Contribute to the world, environment and more : http://www.gridrepublic.org"
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: negative response times again
Posted by monpru <pr...@stud.ntnu.no>.
Hi,
1. I am not able to get that information since we only get to lend servers
for testing per project and have little control over the environment. The
machines are on the same subnet.
2. I think it is milliseconds (if I got your question right).
3. I ran a shorter test with the result tree listener and it looked ok. The
responses in the log also look normal (code 200). This is how it looks like
in the log:
2010/07/07 20:10:49.0189,16,Completion,200,Main scenario 1-3,text,true,0,0
2010/07/07 20:10:49.0282,64,getStatus,200,Main scenario
1-15,text,true,1703,64
2010/07/07 20:10:49.0372,-4507,getStatus,200,Main scenario
1-12,text,true,1703,-4507
Thank you,
Monique
--
View this message in context: http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1045406.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: negative response times again
Posted by Deepak Goel <de...@gmail.com>.
Some thoughts:
1. Can you please post the whole network and server diagram?
2. What is your precision level for measuring latency?
3. It could be possible you might be getting all errors. It would help
if you could post the results also.
Thanks
Deepak
On 7/8/10, monpru <pr...@stud.ntnu.no> wrote:
>
> Hi,
> I got negative latency values in my last load test of a web service. I
> searched the web and found that this might have something to do with the jdk
> I use? I wonder what the problem is or how to ensure that I get correct
> latency in future tests? I am not allowed to use Wireshark or similar to
> check transmission times. The load during the tests was quite low (350
> requests/minute) and none of the machines involved in the test struggled. I
> used 2 thread groups with one Constant Throughput Timer each, all samplers
> are SOAP/XML-RPC Requests. I use csv format for logging.
> I tried to re-run the test using different jdks and got results that
> differed, but had negative and 0 latency in each of the tests.
> I used JMeter 2.3.4 r785646
> and tried with the following
> -Java HotSpot(TM) Server VM (build 1.5.0_16-b02, mixed mode)
> -BEA JRockit(R) (build
> R26.4.0-63_CR302700-72606-1.5.0_06-20061127-1104-linux-ia32
> -Java HotSpot(TM) Server VM (build 1.5.0_06-b05, mixed mode)
> on a machine with CentOS release 4.7 and got negative and 0 latency for all.
>
> I tried on a different machine using BEA WebLogic JRockit(TM) 1.4.2_08 JVM
> R24.5.0-61 on Red Hat Enterprise Linux ES release 4 (Nahant Update 2): here
> I only had 0 load time (the last value) for some requests, but no negative
> or 0 latency. But I am unsure if this means the latency is correct this
> time?
>
> I might be overlooking something obvious here, so I would be thankful for
> hints regarding this problem.
>
> Regards,
> Monique
> --
> View this message in context:
> http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1044999.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
--
Namaskara~Nalama~Guten Tag
--
Keigu
Deepak
+91-9765089593
deicool@gmail.com
Skype: thumsupdeicool
Google talk: deicool
Blog: http://loveandfearless.wordpress.com
Facebook: http://www.facebook.com/deicool
Check out my Work at:
LinkedIn: http://in.linkedin.com/in/thumsupdeicool
"Contribute to the world, environment and more : http://www.gridrepublic.org"
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: negative response times again
Posted by monpru <pr...@stud.ntnu.no>.
Hi,
yes the one I'm having problems with is a virtual machine, I run one JMeter
instance on it for load generation.
Regards,
Monique
--
View this message in context: http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1045409.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: negative response times again
Posted by chaitanya bhatt <bh...@gmail.com>.
Are you by any possibility using Virtual Machines for load injection?
-Chaitanya M Bhatt
http://www.performancecompetence.com
On Thu, Jul 8, 2010 at 5:13 PM, monpru <pr...@stud.ntnu.no> wrote:
>
> Hi,
> I got negative latency values in my last load test of a web service. I
> searched the web and found that this might have something to do with the
> jdk
> I use? I wonder what the problem is or how to ensure that I get correct
> latency in future tests? I am not allowed to use Wireshark or similar to
> check transmission times. The load during the tests was quite low (350
> requests/minute) and none of the machines involved in the test struggled. I
> used 2 thread groups with one Constant Throughput Timer each, all samplers
> are SOAP/XML-RPC Requests. I use csv format for logging.
> I tried to re-run the test using different jdks and got results that
> differed, but had negative and 0 latency in each of the tests.
> I used JMeter 2.3.4 r785646
> and tried with the following
> -Java HotSpot(TM) Server VM (build 1.5.0_16-b02, mixed mode)
> -BEA JRockit(R) (build
> R26.4.0-63_CR302700-72606-1.5.0_06-20061127-1104-linux-ia32
> -Java HotSpot(TM) Server VM (build 1.5.0_06-b05, mixed mode)
> on a machine with CentOS release 4.7 and got negative and 0 latency for
> all.
>
> I tried on a different machine using BEA WebLogic JRockit(TM) 1.4.2_08 JVM
> R24.5.0-61 on Red Hat Enterprise Linux ES release 4 (Nahant Update 2): here
> I only had 0 load time (the last value) for some requests, but no negative
> or 0 latency. But I am unsure if this means the latency is correct this
> time?
>
> I might be overlooking something obvious here, so I would be thankful for
> hints regarding this problem.
>
> Regards,
> Monique
> --
> View this message in context:
> http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1044999.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
Re: negative response times again
Posted by sebb <se...@gmail.com>.
On 8 July 2010 12:43, monpru <pr...@stud.ntnu.no> wrote:
>
> Hi,
> I got negative latency values in my last load test of a web service. I
> searched the web and found that this might have something to do with the jdk
> I use? I wonder what the problem is or how to ensure that I get correct
> latency in future tests? I am not allowed to use Wireshark or similar to
> check transmission times. The load during the tests was quite low (350
> requests/minute) and none of the machines involved in the test struggled. I
> used 2 thread groups with one Constant Throughput Timer each, all samplers
> are SOAP/XML-RPC Requests. I use csv format for logging.
> I tried to re-run the test using different jdks and got results that
> differed, but had negative and 0 latency in each of the tests.
> I used JMeter 2.3.4 r785646
> and tried with the following
> -Java HotSpot(TM) Server VM (build 1.5.0_16-b02, mixed mode)
> -BEA JRockit(R) (build
> R26.4.0-63_CR302700-72606-1.5.0_06-20061127-1104-linux-ia32
> -Java HotSpot(TM) Server VM (build 1.5.0_06-b05, mixed mode)
> on a machine with CentOS release 4.7 and got negative and 0 latency for all.
>
> I tried on a different machine using BEA WebLogic JRockit(TM) 1.4.2_08 JVM
> R24.5.0-61 on Red Hat Enterprise Linux ES release 4 (Nahant Update 2): here
> I only had 0 load time (the last value) for some requests, but no negative
> or 0 latency. But I am unsure if this means the latency is correct this
> time?
Latency of 0 is to be expected sometimes, due to timer resolution.
Also not all samplers support latency, in which case the value will
always be zero.
> I might be overlooking something obvious here, so I would be thankful for
> hints regarding this problem.
>
> Regards,
> Monique
> --
> View this message in context: http://jmeter.512774.n5.nabble.com/negative-response-times-again-tp1044999p1044999.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org