[Nitro] [PATCH] POLS patch for VarChar()

Robert Mela rob at robmela.com
Wed Mar 28 09:02:49 EDT 2007


I'm finding np on the sluggish side.   The issue *might* in part be the 
number of components to download.

There are a lot of metrics that go into assessing a site's performance

 - total start to finish ( from first call to "open" to final byte of 
final asset retrieved )
 - connection initialization ( for single asset, timespan between call 
to "open" and and socket being ready to write )
 - response time ( for single asset, time between last byte of request 
written over socket and first byte of response being ready to read from 
socket )
 - transfer time ( for single asset, time between first response bytes 
arriving on socket and final response bytes arriving on socket )

This is just off the top of my head.  It's been a while since I've done 
performance analysis.

I have to say my world view is challenged by the thought that Internet 
connections in Finland are as bad as the ones in the U.S.  Sigh.  Well, 
adjustment will be minor -- Finland is generally better in most of the 
important metrics :)


Jonathan Buch wrote:
> Hi,
>
>   
>>> I would blame it on np.org not being on a datacenter near Finland :)
>>> The connection times kill the speed, not the page loading at that.
>>>       
>> I find it hard to believe that Greece has a better connection than  
>> Finland ;-)
>> 12 second sounds unrealistic... np.org even uses expire headers and
>> http caching (perhaps I have done something stupid with the apache
>> config on the server though) and ajax for near instant updates.
>>     
>
> I said 'up to', the average is around 3-5 seconds.  Only the first
> request without cache etc is around 12 seconds.
>
> But no, I don't think it's any config error on you part.  It's really
> just looking like long latencies between my connection(s) here and
> np.org.
>
>   
>> anyway ;-)
>>     
>
> Jo
>
>   

-------------- next part --------------
A non-text attachment was scrubbed...
Name: rob.vcf
Type: text/x-vcard
Size: 116 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/nitro-general/attachments/20070328/b6d6a5de/attachment.vcf 


More information about the Nitro-general mailing list