jmeter-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From sebb <seb...@gmail.com>
Subject Re: does jmeter make any difference when run from Different OS with same configuration ?
Date Wed, 12 Jan 2011 09:15:37 GMT
On 12 January 2011 09:08, Mariusz W <mawasak@gmail.com> wrote:
> On Wed, Jan 12, 2011 at 9:29 AM, sebb <sebbaz@gmail.com> wrote:
>
>> On 12 January 2011 08:06, Mariusz W <mawasak@gmail.com> wrote:
>> > Hi,
>> > Two days ago I ran Jmeter (2.4 and 2.3.4) on Amacone EC2 - Windows, Xeon,
>> > Java SE 6 Update 23 (tried 32bit or 64bit) - Retrive all embedded
>> resources
>> > is ON - keep alive on server side is OFF. Elapsed Times (avg, max, min)
>> > where not real - 1 sample, 1 thread showed time 2 or 3 times real elapsed
>> > time (which I measure by stopwatch). I found that every run some css, js,
>> > jpg... has loaded time very high (there were about 80 static content
>> files
>> > and each run other file had that false high load time ). During test
>> outside
>> > ec2 everything is ok. My question is - Do You use Jmeter from cloud
>> > (especially ec2)? Anybody have seen such error?
>>
>> When you run tests outside EC2, was the network route to the target
>> host the same?
>>
>
> No
>
>
>> What do you mean when you say the time is very high on EC2?
>> What is the time otherwise?
>>
>> For example, if EC2 reports a time of 10ms and other time is 1ms, then
>> the ratio is 10 times, but the absolute difference is 9ms, which is
>> similar to timer resolution, and so is not significant.
>>
>>
> From ec2 >100sec (it was shown in Jmeter 2.4)  and >200 sec. on Jmeter
> 2.3.4. - when I check on mobile stopwatch (real time) it was 20 sec (on
> transaction controller) and max 1 sec on that wrong css,jpg, js request on
> server side in accesslog.

That sounds like network problems accessing the server host from EC2.

Try running a simple (short) test accessing a public server such as
www.amazon.com or google from EC2 and see what JMeter times are like
then.

>
>> > Regards,
>> > Mariusz
>> >
>> > On Wed, Jan 12, 2011 at 8:19 AM, Deepak Shetty <shettyd@gmail.com>
>> wrote:
>> >
>> >> In theory it shouldnt.
>> >> In practice the way VMs on different OS's handle threads/sockets/memory
>> may
>> >> impact the runtime behavior (in addition to code bugs , but these should
>> be
>> >> rare).
>> >>
>> >> On Tue, Jan 11, 2011 at 10:45 PM, rajivkumarnandvani <
>> >> rajivkumarnandvani@gmail.com> wrote:
>> >>
>> >> >
>> >> > Hi  All,
>> >> >
>> >> > I know jmeter java based.
>> >> > Does jmeter make any difference when run from Different OS with same
>> >> > configuration?
>> >> >
>> >> > Like run from MAC OS OR windows OS.
>> >> >
>> >> > thanks
>> >> >
>> >> >
>> >> >
>> >> > -----
>> >> >
>> >> > rajivkumarnandvani
>> >> > --
>> >> > View this message in context:
>> >> >
>> >>
>> http://jmeter.512774.n5.nabble.com/does-jmeter-make-any-difference-when-run-from-Different-OS-with-same-configuration-tp3337722p3337722.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
>>
>>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org


Mime
View raw message