Have you considered the IBM jdk v1.1.8?
It is much faster than blackdown.
Our setup is apache-1.3.12, jserv-1.1 on redhat-6.0

> -----Original Message-----
> From: ECS [mailto:[EMAIL PROTECTED]]
> Sent: Friday, March 24, 2000 8:00 PM
> To: ECS
> Subject: ECS Digest #205 - 03/24/00
> 
> 
> ECS Digest #205 - Friday, March 24, 2000
> 
>   Performance Question
>           by "Derek Scherger" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Simon Christian" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Derek Scherger" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Simon Allaway" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Derek Scherger" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Stephan Nagy" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Glen Newton" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "jon *" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "Derek Scherger" <[EMAIL PROTECTED]>
>   Re: Performance Question
>           by "jon *" <[EMAIL PROTECTED]>
> 
> 
> ----------------------------------------------------------------------
> 
> Subject: Performance Question
> From: "Derek Scherger" <[EMAIL PROTECTED]>
> Date: Thu, 23 Mar 2000 23:13:26 -0700
> 
> Hi folks, I've just recently downloaded ECS and started 
> trying to use it
> as a cleaner alternative to what is going to be some really ugly JSP
> code, and my first impressions are very good. It is so much 
> cleaner than
> JAVA embedded in HTML (JSP) or HTML embedded in JAVA
> (out.println("<TAG>") that I'm all but sold. 
> 
> Here's the catch...
> 
> We're generally quite concerned with the performance of our 
> applications
> and new technologies don't generally get the nod unless they compare
> reasonably well with old ones. At the moment, I'm testing ECS against
> some functionally similar JSP code to generate a somewhat large table
> (1000 rows by 10 columns) and JSP is about 10x faster, which 
> is going to
> make ECS a really hard sell.
> 
> First thoughts on this problem were that the number of objects being
> instantiated for this table, on the order of 10,000 or so, 
> was likely a
> big part of the problem. However, a simple test shows that 
> the output or
> actual generation of the HTML takes far longer than instantiating the
> required ECS objects, something like 10x longer generating than
> instantiating. 
> 
> So, first question: does this sound reasonable?
> 
> And, second question: how can I improve the HTML generation time by a
> factor of 10 or so?
> 
> 
> I haven't dug very deep into the code yet to see what's going 
> on and I'm
> hoping someone can shed some light on this. Here's my test program and
> results, I hope they help!
> 
> 
> import java.io.*;
> 
> import org.apache.ecs.*;
> import org.apache.ecs.html.*;
> 
> public class newTimer {
> 
>     public static void main(String args[]) {
>       long start;
>       long end;
>       
>       start = System.currentTimeMillis();
>       System.out.println("Start: " + start);
>       
>       int count = 1000000;
>       
>       for (int i=0; i<count; i++) {
>           String s = new String("foo");
>       }
> 
>       end = System.currentTimeMillis();
>       System.out.println("End:   " + end);
>       
>       System.out.println("Created " + count + " Strings in " 
> + (end - start)
> + "ms");
> 
>       Html page = new Html();
>       Head head = new Head();
>       Body body = new Body();
>       page.addElement(head);
>       page.addElement(body);
> 
>       Table table = new Table();
>       body.addElement(table);
> 
>       start = System.currentTimeMillis();
>       System.out.println("Start: " + start);
> 
>       for (int i=0; i<1000; i++) {
>           TR row = new TR();
>           table.addElement(row);
> 
>           for (int j=0; j<10; j++) {
>               row.addElement(new TD("" + i + "," + j));
>           }
>       }
> 
>       end = System.currentTimeMillis();
>       System.out.println("End:   " + end);
>       
>       System.out.println("Created 1000 row table in " + (end 
> - start) +
> "ms");
> 
>       try {
>           FileOutputStream file = new FileOutputStream("foo.html");
> 
>           start = System.currentTimeMillis();
>           System.out.println("Start: " + start);
>           
>           page.output(file);
>           
>           end = System.currentTimeMillis();
>           System.out.println("End:   " + end);
> 
>           System.out.println("Generated HTML in " + (end - 
> start) + "ms");
>           file.close();
>       }
>       catch (Exception e) {
>           e.printStackTrace();
>       }
> 
>       
>     }
> }
> 
> 
> voodoo$ javac newTimer.java
> voodoo$ java newTimer
> Start: 953877468186
> End:   953877474677
> Created 1000000 Strings in 6491ms
> Start: 953877474730
> End:   953877480225
> Created 1000 row table in 5495ms
> Start: 953877480226
> End:   953877536715
> Generated HTML in 56489ms
> voodoo$ 
> 
> 
> I realize that my timing technique is rather crude but the results are
> consistent.
> 
> 
> The first timing result (6.491 seconds) is a simple benchmark creating
> new Strings, so you might be able to compare numbers from different
> machines. These are from the kaffe-1.0.b4-2 JVM running under Linux
> (RedHat 6.0) on an idle 450 MHz PIII.
> 
> The second timing result (5.495 seconds) shows how long it took to
> instantiate the Html objects representing the page (or table).
> 
> Finally, the third result (56.489 seconds) shows how long it took to
> generate and output the HTML for the table.
> 
> As you can see, ~10x longer to output than to instantiate... :(  Any
> ideas/suggestions/questions?
> 
> 
> 
> Thanks in advance!
> -- 
> Cheers,
> Derek
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Simon Christian" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 09:36:34 +0000
> 
> Derek,
> 
> thought those looked like rather strange results so ran it 
> myself getting these
> timings over 3 runs:
> 
>                  JDK1.2.2        JDK1.1.7A
> Created Strings: 1573            1932
> Created Table:   5194            6823
> Generated HTML:  4158            4850
> 
> Then changed a couple of lines to buffer the output stream 
> (see below).
> 
> Created Strings: 1535            1926
> Created Table:   5156            6813
> Generated HTML:  2353 (-43%)     3518 (-27%)
> 
> This is on a PIII, NT4. jdk1.1.7A required an -mx64M argument 
> to complete the
> test.
> 
> I think you may need to re-evaluate your timing method, and 
> check that other
> factors (e.g. the hard drive?) aren't the bottleneck
> 
> - simon
> 
> changed portion of file:
> 
> 
>         try {
>             FileOutputStream file = new FileOutputStream("foo.html");
>             BufferedOutputStream out = new BufferedOutputStream(file);
> 
>             start = System.currentTimeMillis();
>             // System.out.println("Start: " + start);
> 
>             page.output(out);
> 
>             end = System.currentTimeMillis();
>             // System.out.println("End:   " + end);
> 
>             System.out.println("Generated HTML in " + (end - 
> start) + "ms");
>             out.close();
>             file.close();
>         }
> 
> 
> 
> 
> Derek Scherger wrote:
> 
> > Hi folks, I've just recently downloaded ECS and started 
> trying to use it
> > as a cleaner alternative to what is going to be some really ugly JSP
> > code, and my first impressions are very good. It is so much 
> cleaner than
> > JAVA embedded in HTML (JSP) or HTML embedded in JAVA
> > (out.println("<TAG>") that I'm all but sold.
> >
> > Here's the catch...
> >
> > We're generally quite concerned with the performance of our 
> applications
> > and new technologies don't generally get the nod unless they compare
> > reasonably well with old ones. At the moment, I'm testing 
> ECS against
> > some functionally similar JSP code to generate a somewhat 
> large table
> > (1000 rows by 10 columns) and JSP is about 10x faster, 
> which is going to
> > make ECS a really hard sell.
> >
> > First thoughts on this problem were that the number of objects being
> > instantiated for this table, on the order of 10,000 or so, 
> was likely a
> > big part of the problem. However, a simple test shows that 
> the output or
> > actual generation of the HTML takes far longer than 
> instantiating the
> > required ECS objects, something like 10x longer generating than
> > instantiating.
> >
> > So, first question: does this sound reasonable?
> >
> > And, second question: how can I improve the HTML generation 
> time by a
> > factor of 10 or so?
> >
> > I haven't dug very deep into the code yet to see what's 
> going on and I'm
> > hoping someone can shed some light on this. Here's my test 
> program and
> > results, I hope they help!
> >
> > import java.io.*;
> >
> > import org.apache.ecs.*;
> > import org.apache.ecs.html.*;
> >
> > public class newTimer {
> >
> >     public static void main(String args[]) {
> >         long start;
> >         long end;
> >
> >         start = System.currentTimeMillis();
> >         System.out.println("Start: " + start);
> >
> >         int count = 1000000;
> >
> >         for (int i=0; i<count; i++) {
> >             String s = new String("foo");
> >         }
> >
> >         end = System.currentTimeMillis();
> >         System.out.println("End:   " + end);
> >
> >         System.out.println("Created " + count + " Strings 
> in " + (end - start)
> > + "ms");
> >
> >         Html page = new Html();
> >         Head head = new Head();
> >         Body body = new Body();
> >         page.addElement(head);
> >         page.addElement(body);
> >
> >         Table table = new Table();
> >         body.addElement(table);
> >
> >         start = System.currentTimeMillis();
> >         System.out.println("Start: " + start);
> >
> >         for (int i=0; i<1000; i++) {
> >             TR row = new TR();
> >             table.addElement(row);
> >
> >             for (int j=0; j<10; j++) {
> >                 row.addElement(new TD("" + i + "," + j));
> >             }
> >         }
> >
> >         end = System.currentTimeMillis();
> >         System.out.println("End:   " + end);
> >
> >         System.out.println("Created 1000 row table in " + 
> (end - start) +
> > "ms");
> >
> >         try {
> >             FileOutputStream file = new 
> FileOutputStream("foo.html");
> >
> >             start = System.currentTimeMillis();
> >             System.out.println("Start: " + start);
> >
> >             page.output(file);
> >
> >             end = System.currentTimeMillis();
> >             System.out.println("End:   " + end);
> >
> >             System.out.println("Generated HTML in " + (end 
> - start) + "ms");
> >             file.close();
> >         }
> >         catch (Exception e) {
> >             e.printStackTrace();
> >         }
> >
> >
> >     }
> > }
> >
> > voodoo$ javac newTimer.java
> > voodoo$ java newTimer
> > Start: 953877468186
> > End:   953877474677
> > Created 1000000 Strings in 6491ms
> > Start: 953877474730
> > End:   953877480225
> > Created 1000 row table in 5495ms
> > Start: 953877480226
> > End:   953877536715
> > Generated HTML in 56489ms
> > voodoo$
> >
> > I realize that my timing technique is rather crude but the 
> results are
> > consistent.
> >
> > The first timing result (6.491 seconds) is a simple 
> benchmark creating
> > new Strings, so you might be able to compare numbers from different
> > machines. These are from the kaffe-1.0.b4-2 JVM running under Linux
> > (RedHat 6.0) on an idle 450 MHz PIII.
> >
> > The second timing result (5.495 seconds) shows how long it took to
> > instantiate the Html objects representing the page (or table).
> >
> > Finally, the third result (56.489 seconds) shows how long it took to
> > generate and output the HTML for the table.
> >
> > As you can see, ~10x longer to output than to instantiate... :(  Any
> > ideas/suggestions/questions?
> >
> > Thanks in advance!
> > --
> > Cheers,
> > Derek
> 
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Derek Scherger" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 09:13:58 -0700
> 
> Ok, the kaffe JVM I was using appears to be pretty brutal. Here's the
> results on the same machine with the blackdown 1.1.6 compiler and JVM:
> 
> voodoo$ java newTimer
> Start: 953911983966
> End:   953911988423
> Created 1000000 Strings in 4457ms
> Start: 953911988452
> End:   953911996663
> Created 1000 row table in 8211ms
> Start: 953911996664
> End:   953912000944
> Generated HTML in 4280ms
> 
> Presumably this can be improved a lot with a new JDK too. 
> Here's some more results from a Sun box with JDK 1.2:
> 
> com$ java -version
> java version "1.2.1"
> Solaris VM (build Solaris_JDK_1.2.1_03, native threads, sunwjit)
> 
> com$ java newTimer
> Start: 953911860371
> End: 953911861396
> Created 1000000 Strings in 1025ms
> Start: 953911861546
> End: 953911862781
> Created 1000 row table in 1235ms
> Start: 953911862784
> End: 953911866364
> Walked table in 3580ms
> 
> 
> Anyway, let me try a different approach here. Who else is using ECS
> instead of or in addition to JSP and what are your thoughts on doing
> this? Our impressions are that the code is *much* more 
> readable and the
> design ends up much cleaner using ECS. The only problem we 
> have selling
> ECS internally at this point is that retrieving a similar JSP 
> page is 2x
> to 10x faster than retrieving the same page generated with ECS. On the
> other hand, the HTML generated by ECS is about half the size of the
> corresponding JSP HTML although I'm sure this can be fixed somewhat by
> uglifying the JSP page even further.
> 
> 
> Simon Christian wrote:
> > 
> > Derek,
> > 
> > thought those looked like rather strange results so ran it 
> myself getting these
> > timings over 3 runs:
> > 
> >                  JDK1.2.2        JDK1.1.7A
> > Created Strings: 1573            1932
> > Created Table:   5194            6823
> > Generated HTML:  4158            4850
> > 
> > Then changed a couple of lines to buffer the output stream 
> (see below).
> > 
> > Created Strings: 1535            1926
> > Created Table:   5156            6813
> > Generated HTML:  2353 (-43%)     3518 (-27%)
> > 
> > This is on a PIII, NT4. jdk1.1.7A required an -mx64M 
> argument to complete the
> > test.
> 
> Thanks very much folks!
> -- 
> Cheers,
> Derek
> _____________________________________________________________________
> Derek Scherger                         Echologic Software Corporation
> mailto:[EMAIL PROTECTED]                   http://www.echologic.com
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Simon Allaway" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 10:20:39 -0600
> 
> Derek Scherger wrote:
> > 
> > The only problem we have selling
> > ECS internally at this point is that retrieving a similar 
> JSP page is 2x
> > to 10x faster than retrieving the same page generated with ECS. 
> 
> How would this be related to the HTTP server you're using?
> I've been using Resin 1.1  www.caucho.com
> Their performance figures look excellent on paper, and in practice it
> 'feels' quick.
> 
> Simon
> 
> 
> -- 
> Simon Allaway - University of Chicago
> [EMAIL PROTECTED] -  2-7768
> --
> "I'm against animal testing. They just get all nervous 
> and get the answers wrong."
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Derek Scherger" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 09:38:57 -0700
> 
> We're using Apache 1.3.9 with the GNU JSP engine and Apache 
> JSERV 1.1. 
> 
> Simon Allaway wrote:
> > 
> > Derek Scherger wrote:
> > >
> > > The only problem we have selling
> > > ECS internally at this point is that retrieving a similar 
> JSP page is 2x
> > > to 10x faster than retrieving the same page generated with ECS.
> > 
> > How would this be related to the HTTP server you're using?
> > I've been using Resin 1.1  www.caucho.com
> > Their performance figures look excellent on paper, and in 
> practice it
> > 'feels' quick.
> > 
> > Simon
> > 
> > --
> > Simon Allaway - University of Chicago
> > [EMAIL PROTECTED] -  2-7768
> > --
> > "I'm against animal testing. They just get all nervous
> > and get the answers wrong."
> > 
> > --
> > ------------------------------------------------------------
> > To subscribe:        [EMAIL PROTECTED]
> > To unsubscribe:      [EMAIL PROTECTED]
> > Archives and Other:  <http://java.apache.org/main/mail.html>
> > Problems?:           [EMAIL PROTECTED]
> 
> -- 
> Cheers,
> Derek
> _____________________________________________________________________
> Derek Scherger                         Echologic Software Corporation
> mailto:[EMAIL PROTECTED]                   http://www.echologic.com
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Stephan Nagy" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 10:36:15 -0600
> 
> Derek Scherger wrote:
> 
> > Hi folks, I've just recently downloaded ECS and started 
> trying to use it
> > as a cleaner alternative to what is going to be some really ugly JSP
> > code, and my first impressions are very good. It is so much 
> cleaner than
> > JAVA embedded in HTML (JSP) or HTML embedded in JAVA
> > (out.println("<TAG>") that I'm all but sold.
> >
> > Here's the catch...
> >
> > We're generally quite concerned with the performance of our 
> applications
> > and new technologies don't generally get the nod unless they compare
> > reasonably well with old ones. At the moment, I'm testing 
> ECS against
> > some functionally similar JSP code to generate a somewhat 
> large table
> > (1000 rows by 10 columns) and JSP is about 10x faster, 
> which is going to
> > make ECS a really hard sell.
> >
> 
> Sounds about right.  htmlKona is about 2x to 3x faster then 
> ECS.  I would expect
> jsp to be even faster as it doesn't generate any html from 
> objects, they are
> always just strings.  And doing a println("<sometag here>") 
> is going to be faster
> then calling output or to string on an ecs object.  I would welcome in
> performance improvements.
> 
> >
> > First thoughts on this problem were that the number of objects being
> > instantiated for this table, on the order of 10,000 or so, 
> was likely a
> > big part of the problem.
> 
> modern vm's do this really well.
> 
> > However, a simple test shows that the output or
> > actual generation of the HTML takes far longer than 
> instantiating the
> > required ECS objects, something like 10x longer generating than
> > instantiating.
> >
> 
> Sounds about right modern vm's handle object creation and 
> distruction realitivly
> well.
> 
> >
> > So, first question: does this sound reasonable?
> >
> > And, second question: how can I improve the HTML generation 
> time by a
> > factor of 10 or so?
> >
> 
> Some things to check that will impact performance are 
> filtering & prettyprinting,
> they should be turned off. I don't think you will be able to get a 10x
> improvment.  The ecs core is pretty solid at this point and 
> there aren't any
> glaringly obvious performance bottlenecks. You might get 
> lucky and improve the
> speed by 2x or 3x.  If you are able to get any performance 
> increases from the
> core, send me a patch and I will happily add them.
> 
> >
> > I haven't dug very deep into the code yet to see what's 
> going on and I'm
> > hoping someone can shed some light on this. Here's my test 
> program and
> > results, I hope they help!
> >
> > import java.io.*;
> >
> > import org.apache.ecs.*;
> > import org.apache.ecs.html.*;
> >
> > public class newTimer {
> >
> >     public static void main(String args[]) {
> >         long start;
> >         long end;
> >
> >         start = System.currentTimeMillis();
> >         System.out.println("Start: " + start);
> >
> >         int count = 1000000;
> >
> >         for (int i=0; i<count; i++) {
> >             String s = new String("foo");
> >         }
> >
> >         end = System.currentTimeMillis();
> >         System.out.println("End:   " + end);
> >
> >         System.out.println("Created " + count + " Strings 
> in " + (end - start)
> > + "ms");
> >
> >         Html page = new Html();
> >         Head head = new Head();
> >         Body body = new Body();
> >         page.addElement(head);
> >         page.addElement(body);
> >
> >         Table table = new Table();
> >         body.addElement(table);
> >
> >         start = System.currentTimeMillis();
> >         System.out.println("Start: " + start);
> >
> >         for (int i=0; i<1000; i++) {
> >             TR row = new TR();
> >             table.addElement(row);
> >
> >             for (int j=0; j<10; j++) {
> >                 row.addElement(new TD("" + i + "," + j));
> >             }
> >         }
> >
> >         end = System.currentTimeMillis();
> >         System.out.println("End:   " + end);
> >
> >         System.out.println("Created 1000 row table in " + 
> (end - start) +
> > "ms");
> >
> >         try {
> >             FileOutputStream file = new 
> FileOutputStream("foo.html");
> >
> >             start = System.currentTimeMillis();
> >             System.out.println("Start: " + start);
> >
> >             page.output(file);
> >
> >             end = System.currentTimeMillis();
> >             System.out.println("End:   " + end);
> >
> >             System.out.println("Generated HTML in " + (end 
> - start) + "ms");
> >             file.close();
> >         }
> >         catch (Exception e) {
> >             e.printStackTrace();
> >         }
> >
> >
> >     }
> > }
> >
> > voodoo$ javac newTimer.java
> > voodoo$ java newTimer
> > Start: 953877468186
> > End:   953877474677
> > Created 1000000 Strings in 6491ms
> > Start: 953877474730
> > End:   953877480225
> > Created 1000 row table in 5495ms
> > Start: 953877480226
> > End:   953877536715
> > Generated HTML in 56489ms
> > voodoo$
> >
> > I realize that my timing technique is rather crude but the 
> results are
> > consistent.
> 
> You should buffer you output stream.
> 
> >
> >
> > The first timing result (6.491 seconds) is a simple 
> benchmark creating
> > new Strings, so you might be able to compare numbers from different
> > machines. These are from the kaffe-1.0.b4-2 JVM running under Linux
> > (RedHat 6.0) on an idle 450 MHz PIII.
> 
> The solaris exact vm screams on this stuff.  It will take an 
> ecs document that
> contains somewhere in the neighborhood of 800,000 ecs objects 
> and spit it out in
> approx. 8 to 12 seconds. ( don't have exact numbers been 
> awhile since i
> benchmarked it )
> 
> -stephan
> 
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Glen Newton" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 12:22:48 -0500
> 
> 
> Comments on ECS perfermance:
> 
> Servlets are relatively long-lived and create and destroy numerous
> objects. This is expensive. If you can reduce these numbers, it is a
> Good Thing(tm). Now, instead of always new'ing ECS objects, how about
> caching ecs objects? Use a factory class which manages a cache of ecs
> objects, explicitly calling factory.destroy(ecs_object) which makes
> the object available to the cache once again. It would be useful if
> all ecs objects had some sort of 'init()' method which would reset all
> the values to the defaults (i am behind the times: i am not sure if
> this init() method exists...), as you have to watch out that the
> cached objects do not carry anything unwanted along with them... 
> 
> i might be willing to write this caching class (es?) if 
> someone were to
> add this init method to ecs...
> 
> -glen
> 
> 
> 
> > Derek Scherger wrote:
> > 
> > > Hi folks, I've just recently downloaded ECS and started 
> trying to use it
> > > as a cleaner alternative to what is going to be some 
> really ugly JSP
> > > code, and my first impressions are very good. It is so 
> much cleaner than
> > > JAVA embedded in HTML (JSP) or HTML embedded in JAVA
> > > (out.println("<TAG>") that I'm all but sold.
> > >
> > > Here's the catch...
> > >
> > > We're generally quite concerned with the performance of 
> our applications
> > > and new technologies don't generally get the nod unless 
> they compare
> > > reasonably well with old ones. At the moment, I'm testing 
> ECS against
> > > some functionally similar JSP code to generate a somewhat 
> large table
> > > (1000 rows by 10 columns) and JSP is about 10x faster, 
> which is going to
> > > make ECS a really hard sell.
> > >
> > 
> > Sounds about right.  htmlKona is about 2x to 3x faster then 
> ECS.  I would expect
> > jsp to be even faster as it doesn't generate any html from 
> objects, they are
> > always just strings.  And doing a println("<sometag here>") 
> is going to be faster
> > then calling output or to string on an ecs object.  I would 
> welcome in
> > performance improvements.
> > 
> > >
> > > First thoughts on this problem were that the number of 
> objects being
> > > instantiated for this table, on the order of 10,000 or 
> so, was likely a
> > > big part of the problem.
> > 
> > modern vm's do this really well.
> > 
> > > However, a simple test shows that the output or
> > > actual generation of the HTML takes far longer than 
> instantiating the
> > > required ECS objects, something like 10x longer generating than
> > > instantiating.
> > >
> > 
> > Sounds about right modern vm's handle object creation and 
> distruction realitivly
> > well.
> > 
> > >
> > > So, first question: does this sound reasonable?
> > >
> > > And, second question: how can I improve the HTML 
> generation time by a
> > > factor of 10 or so?
> > >
> > 
> > Some things to check that will impact performance are 
> filtering & prettyprinting,
> > they should be turned off. I don't think you will be able 
> to get a 10x
> > improvment.  The ecs core is pretty solid at this point and 
> there aren't any
> > glaringly obvious performance bottlenecks. You might get 
> lucky and improve the
> > speed by 2x or 3x.  If you are able to get any performance 
> increases from the
> > core, send me a patch and I will happily add them.
> > 
> 
> stuff deleted......
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "jon *" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 09:17:42 -0800
> 
> on 3/24/00 8:13 AM, Derek Scherger <[EMAIL PROTECTED]> wrote:
> 
> > Anyway, let me try a different approach here. Who else is using ECS
> > instead of or in addition to JSP and what are your thoughts on doing
> > this? Our impressions are that the code is *much* more 
> readable and the
> > design ends up much cleaner using ECS. The only problem we 
> have selling
> > ECS internally at this point is that retrieving a similar 
> JSP page is 2x
> > to 10x faster than retrieving the same page generated with 
> ECS. On the
> > other hand, the HTML generated by ECS is about half the size of the
> > corresponding JSP HTML although I'm sure this can be fixed 
> somewhat by
> > uglifying the JSP page even further.
> 
> The issue with speed is obviously an issue with JVM speed. 
> ECS does have an
> overhead, but it isn't so much that it outweighs its advantages.
> 
> I personally wouldn't use ECS as a replacement for JSP. I 
> also wouldn't use
> JSP. ;-) So, what would I use today? I would use a 
> combination of Turbine
> and WebMacro or Freemarker or I would use Cocoon.
> 
> -jon
> 
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "Derek Scherger" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 12:48:02 -0700
> 
> Thanks for your reply Stephan, I have a couple follow up questions for
> you...
> 
> > Some things to check that will impact performance are 
> filtering & prettyprinting,
> > they should be turned off. I don't think you will be able 
> to get a 10x
> > improvment.  The ecs core is pretty solid at this point and 
> there aren't any
> > glaringly obvious performance bottlenecks. You might get 
> lucky and improve the
> > speed by 2x or 3x.  If you are able to get any performance 
> increases from the
> > core, send me a patch and I will happily add them.
> 
> filter_state=false
> filter_attrivute_state=false
> pretty_print=false
> 
> I'll let you know if we do manage to speed things up anywhere. The
> numbers from my 1000 row by 10 column table are all over the map in
> terms of object creation verses html generation times depending on the
> VM and machine being used. Initially I thought that the html 
> generation
> was bad but I now have other cases where it's the same or faster than
> object creation.
> 
> > The solaris exact vm screams on this stuff.  It will take 
> an ecs document that
> > contains somewhere in the neighborhood of 800,000 ecs 
> objects and spit it out in
> > approx. 8 to 12 seconds. ( don't have exact numbers been 
> awhile since i
> > benchmarked it )
> 
> I've never heard of the solaris "exact" vm... what/where is it? 
> 
> Say you get 800,000 ecs objects in 10 seconds, that's 100,000 objects
> per second which sounds pretty good all right, what type of 
> machine did
> you get these numbers on? 
> i.e. how big and fast was it? :)
> 
> 
> Thanks again
> -- 
> Cheers,
> Derek
> _____________________________________________________________________
> Derek Scherger                         Echologic Software Corporation
> mailto:[EMAIL PROTECTED]                   http://www.echologic.com
> 
> ----------------------------------------------------------------------
> 
> Subject: Re: Performance Question
> From: "jon *" <[EMAIL PROTECTED]>
> Date: Fri, 24 Mar 2000 12:04:38 -0800
> 
> on 3/24/00 11:48 AM, Derek Scherger <[EMAIL PROTECTED]> wrote:
> 
> > I've never heard of the solaris "exact" vm... what/where is it?
> 
> Java 2 Standard Edition (1.2.2_05)
> http://www.sun.com/software/solaris/java/download.html
> 
> by far one of the fastest JVM's out there...they call it the exact vm
> because of the type of gc that it uses...exact gc...
> 
> you should also make sure to add in HotSpot as well...;-)
> 
> http://java.sun.com/products/hotspot/index.html
> 
> it is 30% faster than the 1.0 version of hotspot...
> 
> all this speed stuff is by far a factor of the JVM that you 
> are using...
> 
> -jon
> 
> 
> ----------------------------------------------------------------------
> End of ECS Digest
> 
> 
> --
> ------------------------------------------------------------
> To subscribe:        [EMAIL PROTECTED]
> To unsubscribe:      [EMAIL PROTECTED]
> Archives and Other:  <http://java.apache.org/main/mail.html>
> Problems?:           [EMAIL PROTECTED]
> 


--
------------------------------------------------------------
To subscribe:        [EMAIL PROTECTED]
To unsubscribe:      [EMAIL PROTECTED]
Archives and Other:  <http://java.apache.org/main/mail.html>
Problems?:           [EMAIL PROTECTED]

Reply via email to