Jump to content
  • 0

http response time caused by app?


steffenmand

Question

Hi i just tried .670 yesterday where i noticed that i got 5 ms response time on chunks. however uploads would fail at a point due to a number having to be positive.

 

In the new .672 this has been fixed, but my response time from .647 - .672 has growth from 4000ms to 10000ms. as .670 showed EXTREME  fast response times i was wondering if you are doing stuff which delays the response? Back in the really early days i started with 400ms which now have increased to 10000ms giving less speed due to wait time.

 

Is there anything that can be done to eliminate all this extra response time as .670 obviously is bypassing some stuff and giving 5ms?

Link to comment
Share on other sites

11 answers to this question

Recommended Posts

  • 0

These builds are alpha builds.   They are heavily in flux, because Alex is doing a lot of testing.

 

Once there is a numbered build, it should be "set", and if that build has an issue, let us know.

 

(this is part of the reason for the "second numbered build" recommendation for you guys.... these should be more stable, and less likely to have weird issues like this). 

 

IIRC, the reason for this is that there are a lot of error checking code that has been "flipped on" 

Link to comment
Share on other sites

  • 0

I understand that they are alpha builds :-) just noticed the huge change in response time which is a bottleneck in performance. If the response time was as good as i saw, then 20 mb would fully utilize my bandwidth and you could fully ignore the bigger chunks request as 1 gbit would be completely used with such a fast responsetime. however glad to hear that it will get better for release :-).

Link to comment
Share on other sites

  • 0

I have noticed with build .680 that my response times are between 2,500ms and 6,500ms. I dont know how much of an effect this has on things but I do know that with some of the older builds i was getting response times around 100ms.

 

 

Got the same thing but as he says it can be a lot of debugging stuff.

 

With the version that failed i got 5ms response times, so they can get really fast when they start to optimize for that. i got 3-4x the speeds with that low response time so it will make a difference

Link to comment
Share on other sites

  • 0

I have noticed with build .680 that my response times are between 2,500ms and 6,500ms. I dont know how much of an effect this has on things but I do know that with some of the older builds i was getting response times around 100ms.

 

 

Got the same thing but as he says it can be a lot of debugging stuff.

 

With the version that failed i got 5ms response times, so they can get really fast when they start to optimize for that. i got 3-4x the speeds with that low response time so it will make a difference

 

 

 

 

Is this still happening in the 1.0.0.682 build?

Link to comment
Share on other sites

  • 0

Also, which providers is this for? 

 

Also, it may be best to capture the network traffic using "FiddlerCap" 

http://www.telerik.com/fiddler/fiddlercap

(you'll want to enable the HTTPS option, and install the root CA certificate when it prompts). 

 

And probably enable logging. 

 

 

That should get a detailed look at what is going on.

Link to comment
Share on other sites

  • 0

using the version listed in the OP i got like 5 ms from Google. Over time the application just got a slower and slower response time and its not due to something with the network. 

 

It is most definately something in Cloud Drive causing the extra time.

 

when i first tried google  i got 400ms, now it sometimes peak above 12000ms.

 

Jsing .670 response times are perfect - most likely because some code is disabled here, i went from 300 mbit download to 700-800 mbit on the same chunk size  - too bad the release were flawed :-D

Link to comment
Share on other sites

  • 0

If your'e using builds prior to 680, there was a bug that would cause the excessively long timeouts.  In this case, try the 682 build, and see if that helps. 

 

Otherwise, check the cluster size for the CloudDrive disk (you can do that buy running "fsutil fsinfo ntfsinfo x:".  Check the "Bytes per Cluster" size. 

 

If it's using 64kb per cluster, it looks like there is some weird performance issues with using the larger cluster size.  

 

 

Otherwise, the "fiddlerCap" thing will capture the web traffic, so we can take a look at what is going on there (including listing the response times), and the normal logging may indicate other issues, if that is the case. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Answer this question...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...