HTTP Acceleration

119 views
Skip to first unread message

josephanu

unread,
Jun 15, 2015, 9:51:58 AM6/15/15
to adf-met...@googlegroups.com
Hi,

Eager to hear members of the ADF EMG thoughts and experiences about how to handle performance issues relates to HTTP side of the ADF application. Sometimes external end users who accesses the application over the internet sees performance issues like initial page or each page load with a fresh cache is very long, or it takes long time for response to each request to initiate. Our internal stress testing performed with 200 concurrent users found that average request processing time is always less than 3 seconds. None of the internal users are seeing any performance issue with the application when external users faces slowness in application response.

We are using the Oracle Webtier (Oracle HTTP Server) as weblogic proxy (mod_wls_ohs) for load balancing and here OHS set content type to GZIP/deflate to compress/decompress all contents so our application doesn't have to think about the data being compressed at all.

Returning to the original question, how we can achieve HTTP acceleration to provide to remote users the best response time when they access applications over the internet.

1. Any suggestions on using CDNs like AKAMAI 
2.Any specific further optimization suggestions for applications using over the internet.
3. Performance Enhancing Proxies?

Thanks,
Joseph

Florin Marcus

unread,
Jun 15, 2015, 10:05:26 AM6/15/15
to adf-met...@googlegroups.com
How exactly do you determine that the problem comes from HTTP trafic alone, so there is no relation with server processing time?
I am just trying to understand if you are looking in the right direction.

--
--
You received this message because you are subscribed to the ADF Enterprise Methodology Group (http://groups.google.com/group/adf-methodology). To unsubscribe send email to adf-methodolo...@googlegroups.com
 
All content to the ADF EMG lies under the Creative Commons Attribution 3.0 Unported License (http://creativecommons.org/licenses/by/3.0/). Any content sourced must be attributed back to the ADF EMG with a link to the Google Group (http://groups.google.com/group/adf-methodology).

---
You received this message because you are subscribed to the Google Groups "ADF Enterprise Methodology Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to adf-methodolo...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Aaron Rapp

unread,
Jun 15, 2015, 10:27:41 AM6/15/15
to adf-met...@googlegroups.com
Assuming you don't have any network issues and have checked your requests to make sure there are no 'no-cache' headers. You may want to try using web cache in front of OHS. If you do use web cache you'll need to take some time in configuring your rules to suit. 


--

Mos'ab Abolila

unread,
Jun 16, 2015, 8:11:03 AM6/16/15
to adf-met...@googlegroups.com
I think you can use a hardware load balancing proxy its amazing with a good result , you said that you have a problem using load balancing but you specify that its about HTTP traffic?? are you sure? I think there is a problem in weblogic proxy it self.
good luck.

John Flack

unread,
Jun 16, 2015, 8:38:53 AM6/16/15
to adf-met...@googlegroups.com, jose...@gmail.com
Of course, there are certain tuning parameters in web.xml that you might have with one setting for development, but another for production.  For instance

org.apache.myfaces.trinidad.COMPRESS_VIEW_STATE might be false during development, especially if you are working on a skin, so that you can see the original style class names, but you want it true for production.


See the Performance Guide at http://docs.oracle.com/middleware/1213/core/ASPER/adf.htm#ASPER99129 for details.

Blake Sullivan

unread,
Jun 19, 2015, 12:18:30 AM6/19/15
to adf-met...@googlegroups.com, jose...@gmail.com
On Jun 16, 2015, at 5:38 AM, John Flack <Jo...@smdi.com> wrote:

Of course, there are certain tuning parameters in web.xml that you might have with one setting for development, but another for production.  For instance

org.apache.myfaces.trinidad.COMPRESS_VIEW_STATE might be false during development, especially if you are working on a skin, so that you can see the original style class names, but you want it true for production.


I believe that you meant “org.apache.myfaces.trinidad.DISABLE_CONTENT_COMPRESSION” here. “org.apache.myfaces.trinidad.COMPRESS_VIEW_STATE” takes the server view state and then Serializes and compresses it so that it takes up less live memory.

— Blake Sullivan


See the Performance Guide at http://docs.oracle.com/middleware/1213/core/ASPER/adf.htm#ASPER99129 for details.


On Monday, June 15, 2015 at 9:51:58 AM UTC-4, josephanu wrote:
Hi,

Eager to hear members of the ADF EMG thoughts and experiences about how to handle performance issues relates to HTTP side of the ADF application. Sometimes external end users who accesses the application over the internet sees performance issues like initial page or each page load with a fresh cache is very long, or it takes long time for response to each request to initiate. Our internal stress testing performed with 200 concurrent users found that average request processing time is always less than 3 seconds. None of the internal users are seeing any performance issue with the application when external users faces slowness in application response.

We are using the Oracle Webtier (Oracle HTTP Server) as weblogic proxy (mod_wls_ohs) for load balancing and here OHS set content type to GZIP/deflate to compress/decompress all contents so our application doesn't have to think about the data being compressed at all.

Returning to the original question, how we can achieve HTTP acceleration to provide to remote users the best response time when they access applications over the internet.

1. Any suggestions on using CDNs like AKAMAI 
2.Any specific further optimization suggestions for applications using over the internet.
3. Performance Enhancing Proxies?

Thanks,
Joseph

John Flack

unread,
Jun 19, 2015, 8:12:43 AM6/19/15
to adf-met...@googlegroups.com

You are right, Blake, I meant “org.apache.myfaces.trinidad.DISABLE_CONTENT_COMPRESSION.  Oops.

 

The main point is that if you are sure that the performance problem is client side, rather than server side, you want to look at reducing the amount of data being sent to the client.  One of the ways of doing this is in the context initialization parameters – and Blake mentioned the one I meant.  Another way is to take a look at your page layouts.  For instance, I’m now doing with af:panelGridLayout some of the things that I used to do with multiple nested panelGroups, and it has significantly reduced the amount of HTML it sends to the client.

 

From: adf-met...@googlegroups.com [mailto:adf-met...@googlegroups.com] On Behalf Of Blake Sullivan
Sent: Friday, June 19, 2015 12:18 AM
To: adf-met...@googlegroups.com
Cc: jose...@gmail.com
Subject: Re: [ADF EMG] HTTP Acceleration

 

 

On Jun 16, 2015, at 5:38 AM, John Flack <Jo...@smdi.com> wrote:

 

Of course, there are certain tuning parameters in web.xml that you might have with one setting for development, but another for production.  For instance

org.apache.myfaces.trinidad.COMPRESS_VIEW_STATE might be false during development, especially if you are working on a skin, so that you can see the original style class names, but you want it true for production.

Jeevan G Joseph

unread,
Jun 22, 2015, 6:54:50 AM6/22/15
to adf-met...@googlegroups.com
Joseph,

Apart from the suggestions below, also take a look at basics : page weight, the number of http requests required to render a page -  particularly the first page , since that would trigger a lot of downloads for the js and CSS assets. The general rules I would recommend are :

1. See if re-partitioning your js would save either bytes-on-the-wire or http round-trips for you. 
2. Use the skin editor to remove unused selectors - quite risky, because you lose the ability to use these selectors as you are manually removing them. Development of your pages will have to reflect this choice and discipline .
3. CDN for caching all static assets. And gzipping content where it makes sense ( text/HTML , js , CSS , Json and other character types) gzipping image/png for instance is just gonna drive up the CPU utilization and load average on the Ohs box.

Take a look at these ( although we wrote these for webcenter, much of it is applicable for adf applications in general ): 
--
--
You received this message because you are subscribed to the ADF Enterprise Methodology Group (http://groups.google.com/group/adf-methodology). To unsubscribe send email to adf-methodolo...@googlegroups.com
 
All content to the ADF EMG lies under the Creative Commons Attribution 3.0 Unported License (http://creativecommons.org/licenses/by/3.0/). Any content sourced must be attributed back to the ADF EMG with a link to the Google Group (http://groups.google.com/group/adf-methodology).

---
You received this message because you are subscribed to the Google Groups "ADF Enterprise Methodology Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to adf-methodolo...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--
Sent from Gmail Mobile
Reply all
Reply to author
Forward
0 new messages