On 16-Apr-2002 Alexander Hartmaier wrote:
> I tried for one week to change apache settings and other things but the thing
> is damn slow! Almost 1 minute for the biggest table which has ~2100 lines
> with 16 fields per line.
> The httpd process on the "new" machine uses all available processor
>
Wim
> code. Inside that eval block, the entire results of the query are being
> loaded into memory. Instead of doing that, only put the $dbh->prepare
> and $sth->execute in the eval. In the display code, process each row one at
> a time, via something like this:
That confirms my suspicion - it
On 16-Apr-2002 David Lloyd wrote:
>
> Alexander
>
>> The "old" machine:
>> 800Mhz PIII with 256MB ram
>> RedHat 6.2
>> Apache 1.3.19 with mod_perl 1.25 and embperl 1.3.1
>>
>> The "new" machine:
>> 866Mhz PIII with 256MB ram
>> RedHat 7.2
>> Apache 1.3.22 with mod_perl 1.26 and embperl 2.0b7
>
On 16-Apr-2002 David Lloyd wrote:
>
> Wim
>
>> code. Inside that eval block, the entire results of the query are being
>> loaded into memory. Instead of doing that, only put the $dbh->prepare
>> and $sth->execute in the eval. In the display code, process each row one at
>> a time, via something
It seems to me (from document fragments and example code) that the 'XML
integration' has mostly targetted the XSLT transformations.
What if I just want to parse some XML files?
- are the libxml2 documented functions just bootstrapped into the embperl
namespace? (haven't tested this at all,
Hi!
First thanks for the lot's and quick answers!
@ lloy0076: I don't use postgresql. The database is a dual xeon-500 with 1GB ram
running Oracle 9.0.1i.
@wim: I'll try to change the code of the .epl...i'm very new to embperl (2
weeks) and just starting to learn it.
The filesystem is ext3 for
Alexander
> Do you really think the kernel is the problem?
It could be -- try using the Kernel RPMS created by RedHat and see if
there are any updates on www.redhat.com.
> Is there any way how to debug the script (or the kernel)?
perldoc HTML::Embperl
...and scroll down to debug. I wonder w
On 16-Apr-2002 Alexander Hartmaier wrote:
>
> @wim: I'll try to change the code of the .epl...i'm very new to embperl (2
> weeks) and just starting to learn it.
No problem. Tactics for optimizing perl database code in normal Perl can be
applied to Embperl as well... have a look at:
http://www.s
@lloy0076
>> Do you really think the kernel is the problem?
>It could be -- try using the Kernel RPMS created by RedHat and see if
>there are any updates on www.redhat.com.
I use the newest rpm found on the ftp server (as well as on the homepage).
kernel-2.4.9-31
@wim
I just reworte view-ta
On 16-Apr-2002 David Lloyd wrote:
>> Do you really think the kernel is the problem?
>
> It could be -- try using the Kernel RPMS created by RedHat and see if
> there are any updates on www.redhat.com.
One of the first things I do when setting up any server, is download the source
to latest sta
On 16-Apr-2002 Alexander Hartmaier wrote:
> I just reworte view-table.epl...it isn't faster than before
>
> Here is the new version (the display of the number of lines isn't workling
> any more :(
> Maybe you can tell me what i do false:
Hmmm...
1. I'm unsure whether it's good to use eval
On 16-Apr-2002 Alexander Hartmaier wrote:
> I just reworte view-table.epl...it isn't faster than before
>
> Here is the new version (the display of the number of lines isn't workling
> any more :(
> Maybe you can tell me what i do false:
Hmmm...
1. I'm unsure whether it's good to use ev
On 16-Apr-2002 Cameron McBride wrote:
>> In a nutshell, create a hidden layer. In that hidden layer, do a
>> GET/POST to a embperl/mod_perl page. The request will get sent to the
>> hidden in-memory document, which can then be parsed back into javascript
>> variables.
>
> ooo, perhaps there migh
I just checked everything with a co-worker who maintains this web-app longer
than i do :).
The downgrad from apache-1.3.22-2 to apache-1.3.20-16 hasn't helped!
Now I will update my kernel to 2.4.18. (no rpmcompile by myself :)
@wim: The URL you wrote for DBI optimization was a great help!
> If a Javascript enabled browser is already a requirement, it's a great way of
> turning a browser from a stateless machine into a stateful one. For example,
You have a very good point here, but is it a realistic one? (stateful
thing) Almost every time I have played with things in Javascript --
At 3:43 AM -0700 4/16/02, Wim Kerkhoff wrote:
>If a Javascript enabled browser is already a requirement, it's a great way of
You mean, "If a browser supporting layers is already requirement".
And of course you need to check it against
Netscape/Mozilla/IEWin/IEMac/Opera/
Mind you, I don't
>I just checked everything with a co-worker who maintains this web-app longer
>than i do :).
>
>The downgrad from apache-1.3.22-2 to apache-1.3.20-16 hasn't helped!
>
>Now I will update my kernel to 2.4.18. (no rpmcompile by myself :)
>
The kernel update was unnecessary cause it brought no
On 16-Apr-2002 Kee Hinckley wrote:
> At 3:43 AM -0700 4/16/02, Wim Kerkhoff wrote:
>>If a Javascript enabled browser is already a requirement, it's a great way of
>
> You mean, "If a browser supporting layers is already requirement".
> And of course you need to check it against
> Netscape/Moz
On 16-Apr-2002 Alexander Hartmaier wrote:
>
>
> I just checked everything with a co-worker who maintains this web-app longer
> than i do :).
>
> The downgrad from apache-1.3.22-2 to apache-1.3.20-16 hasn't helped!
>
> Now I will update my kernel to 2.4.18. (no rpmcompile by myself :)
>
>
On 16-Apr-2002 Cameron McBride wrote:
>> If a Javascript enabled browser is already a requirement, it's a great way
>> of
>> turning a browser from a stateless machine into a stateful one. For example,
>
> You have a very good point here, but is it a realistic one? (stateful
> thing) Almost eve
At 2:23 PM -0700 4/16/02, Wim Kerkhoff wrote:
>With JavaScript form validation, it's pretty easy to allow browsers with no JS
>support, or JS turned off, to still be able to submit the form without writing
>extra code. The JS code just gets ignored. However, with the XML-RPC idea,
You should alwa
On 16-Apr-2002 Alexander Hartmaier wrote:
>
> What does eval do??? (I'm really new!).
run 'perldoc -f eval' to get an explanation, or see your favorite perl book :-)
> What has it to look like if I don't use eval around the perl code at the top
> of the script?
Just remove the 'eval {' at the
On 16-Apr-2002 Kee Hinckley wrote:
> At 2:23 PM -0700 4/16/02, Wim Kerkhoff wrote:
>
> You should always be doing validation on the back-end as well, if
> only to defend against malcious users.
That goes without saying... protection against both malicious and brain
dead users.
>>I found Embp
On 16-Apr-2002 Cameron McBride wrote:
> It seems to me (from document fragments and example code) that the 'XML
> integration' has mostly targetted the XSLT transformations.
>
> What if I just want to parse some XML files?
>
> - are the libxml2 documented functions just bootstrapped into the
At 3:19 PM -0700 4/16/02, Wim Kerkhoff wrote:
>Where does one start? Are there any installers for other CGI
>projects that can
>be imitated? I started writing an installer for acmemail [1] in the style of
I don't know what's out there for CGI-type installations.
I'd think the first thing would
On 16-Apr-2002 Kee Hinckley wrote:
> At 3:19 PM -0700 4/16/02, Wim Kerkhoff wrote:
>>Where does one start? Are there any installers for other CGI
>>projects that can
>>be imitated?
>
> I don't know what's out there for CGI-type installations.
> I'd think the first thing would be to see if can
Hi,
> > That would be great! While I don't think it's a kernel isusse, it may
have
> > to do with the Perl version.
>
> Ok, did it. The mod_info part is the same as before (ie pretty much
> blank) -- but the directives inside location blocks work. ;-)
I don't have any idea why mod_info fails (
>
> What I had in mind for Embperl (using the Embperl::Object stuff) was an
> inheritance tree for the request object. But I want my base.epl to be
> in the middle:
> (the order is via my understanding of precedence, most -> least)
>
> REQUESTED PAGE -->base.epl--> defaults.epl
> (the
>Now my co-worker and I changed the epl. It seems that the large amount of
while
>loops (2100 x 16 = 33.600) is a problem in Embperl 2.0b7.
I guess you are right. There is an hash table to handle iteration inside of
Embperl 2.0, which seems to scale not so well.
To verfiy this I have increased
On Mon, 15 Apr 2002, Wim Kerkhoff wrote:
> In a nutshell, create a hidden layer. In that hidden layer, do a
> GET/POST to a embperl/mod_perl page. The request will get sent to the
> hidden in-memory document, which can then be parsed back into javascript
> variables.
That's a security hole! I d
> It seems to me (from document fragments and example code) that the 'XML
> integration' has mostly targetted the XSLT transformations.
>
It's maily intended for transforming it and generating output. For now this
means mainly XSLT.
> What if I just want to parse some XML files?
>
> - are the
On Tue, 16 Apr 2002, Wim Kerkhoff wrote:
> On 16-Apr-2002 Alexander Hartmaier wrote:
>> I just reworte view-table.epl...it isn't faster than before
>>
>> Here is the new version (the display of the number of lines isn't workling
>> any more :(
>> Maybe you can tell me what i do false:
>
> Hmm
Hi,
great that it works now! Please answer to the list also, other people may
have the same problem and are happy when they find the answer in the mailing
list archives
Gerald
> Hallo,
>
> > > What I had in mind for Embperl (using the Embperl::Object stuff) was
an
> > > inheritance tree for
> >
> > 1. I'm unsure whether it's good to use eval around so much code... is
> > it the most efficient method? You could wrap it just around the
> > ->prepare and ->execute. Actually, you don't eval at all. The critical
> > DBI method calls return true or false, which you can catch:
> >
There a
Hallo,
> > What I had in mind for Embperl (using the Embperl::Object stuff) was an
> > inheritance tree for the request object. But I want my base.epl to be
> > in the middle:
> > (the order is via my understanding of precedence, most -> least)
> >
> > REQUESTED PAGE -->base.epl--> de
> > It seems to me (from document fragments and example code) that the 'XML
> > integration' has mostly targetted the XSLT transformations.
> >
>
> It's maily intended for transforming it and generating output. For now this
> means mainly XSLT.
Ahh, perhaps I am starting to get it. Basically th
>
> Ahh, perhaps I am starting to get it. Basically the XSLT recipes
> configure the processing of the code to use XML libraries, and when you
> do an Execute on them, they make the calls using the specified XML
> library (libxml2 in my case). So all the XSLT functionality is done
> by the respe
37 matches
Mail list logo
