Mailing List Archive

If any of the mod_php extensions are not thread-safe, you will need to use
the prefork mpm, which will indeed bloat every httpd worker. This is not
the ideal nor recommended configuration.

Instead, use the event mpm and proxy_fcgi to pass the request to php-fpm.

Alternatively, you can recompile php to be thread-safe and use event mpm
with mod_php, which will give you a small execution speed advantage.

The statement you posted is more or less FUD which leaves out very
important details.

On Sat, Sep 30, 2023 at 2:56?AM Jason Long <hack3rcon@yahoo.com.invalid>
wrote:

> Hello,
> Is the following sentence correct?
> "The way Apache loads PHP in its standard setup (with mod_php) compared to
> Nginx alone puts it at a disadvantage. You will see performance gains,
> particularly in memory usage, just by switching to Nginx, given you're
> using a PHP-driven application."
>
> Thank you.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> For additional commands, e-mail: users-help@httpd.apache.org
>
>

Additionally, your recent string of questions to this mailing list come off
as a bit trollish.

On Sat, Sep 30, 2023 at 3:04?AM Frank Gingras <thumbs@apache.org> wrote:

> If any of the mod_php extensions are not thread-safe, you will need to use
> the prefork mpm, which will indeed bloat every httpd worker. This is not
> the ideal nor recommended configuration.
>
> Instead, use the event mpm and proxy_fcgi to pass the request to php-fpm.
>
> Alternatively, you can recompile php to be thread-safe and use event mpm
> with mod_php, which will give you a small execution speed advantage.
>
> The statement you posted is more or less FUD which leaves out very
> important details.
>
> On Sat, Sep 30, 2023 at 2:56?AM Jason Long <hack3rcon@yahoo.com.invalid>
> wrote:
>
>> Hello,
>> Is the following sentence correct?
>> "The way Apache loads PHP in its standard setup (with mod_php) compared
>> to Nginx alone puts it at a disadvantage. You will see performance gains,
>> particularly in memory usage, just by switching to Nginx, given you're
>> using a PHP-driven application."
>>
>> Thank you.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
>> For additional commands, e-mail: users-help@httpd.apache.org
>>
>>

Hello,
Thank you so much for your info.
Why are they trollish? I am curious to learn more.
Is there a fair comparison between Apache and Nginx?



On Saturday, September 30, 2023 at 10:35:12 AM GMT+3:30, Frank Gingras <thumbs@apache.org> wrote:





Additionally, your recent string of questions to this mailing list come off as a bit trollish.

On Sat, Sep 30, 2023 at 3:04?AM Frank Gingras <thumbs@apache.org> wrote:
> If any of the mod_php extensions are not thread-safe, you will need to use the prefork mpm, which will indeed bloat every httpd worker. This is not the ideal nor recommended configuration.
>
> Instead, use the event mpm and proxy_fcgi to pass the request to php-fpm.
>
> Alternatively, you can recompile php to be thread-safe and use event mpm with mod_php, which will give you a small execution speed advantage.
>
> The statement you posted is more or less FUD which leaves out very important details.
>
> On Sat, Sep 30, 2023 at 2:56?AM Jason Long <hack3rcon@yahoo.com.invalid> wrote:
>> Hello,
>> Is the following sentence correct?
>> "The way Apache loads PHP in its standard setup (with mod_php) compared to Nginx alone puts it at a disadvantage. You will see performance gains, particularly in memory usage, just by switching to Nginx, given you're using a PHP-driven application."
>>
>> Thank you.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
>> For additional commands, e-mail: users-help@httpd.apache.org
>>
>>
>


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org

There might be some online, however, due to the different architectures,
they are not likely to be terribly useful. Do avoid the ones that bash
needlessly either product.

On Sat, Sep 30, 2023 at 3:09?AM Jason Long <hack3rcon@yahoo.com.invalid>
wrote:

> Hello,
> Thank you so much for your info.
> Why are they trollish? I am curious to learn more.
> Is there a fair comparison between Apache and Nginx?
>
>
>
> On Saturday, September 30, 2023 at 10:35:12 AM GMT+3:30, Frank Gingras <
> thumbs@apache.org> wrote:
>
>
>
>
>
> Additionally, your recent string of questions to this mailing list come
> off as a bit trollish.
>
> On Sat, Sep 30, 2023 at 3:04?AM Frank Gingras <thumbs@apache.org> wrote:
> > If any of the mod_php extensions are not thread-safe, you will need to
> use the prefork mpm, which will indeed bloat every httpd worker. This is
> not the ideal nor recommended configuration.
> >
> > Instead, use the event mpm and proxy_fcgi to pass the request to php-fpm.
> >
> > Alternatively, you can recompile php to be thread-safe and use event mpm
> with mod_php, which will give you a small execution speed advantage.
> >
> > The statement you posted is more or less FUD which leaves out very
> important details.
> >
> > On Sat, Sep 30, 2023 at 2:56?AM Jason Long <hack3rcon@yahoo.com.invalid>
> wrote:
> >> Hello,
> >> Is the following sentence correct?
> >> "The way Apache loads PHP in its standard setup (with mod_php) compared
> to Nginx alone puts it at a disadvantage. You will see performance gains,
> particularly in memory usage, just by switching to Nginx, given you're
> using a PHP-driven application."
> >>
> >> Thank you.
> >>
> >> ---------------------------------------------------------------------
> >> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> >> For additional commands, e-mail: users-help@httpd.apache.org
> >>
> >>
> >
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> For additional commands, e-mail: users-help@httpd.apache.org
>
>

>
>> Why are they trollish? I am curious to learn more.
>>
>>
Perhaps because you asked the same question in 2021, you got a very
constructive response and still you are asking the same thing two years
later...

--
Daniel Ferradal
HTTPD Project
#httpd help at Libera.Chat

On Sat, 30 Sept 2023, 12:26 Jason Long, <hack3rcon@yahoo.com.invalid> wrote:

> Hello,
> Is the following sentence correct?
> "The way Apache loads PHP in its standard setup (with mod_php) compared to
> Nginx alone puts it at a disadvantage. You will see performance gains,
> particularly in memory usage, just by switching to Nginx, given you're
> using a PHP-driven application."
>

You will have to performance benchmark both Apache, Nginx to make a
statement.

We use Apache with PHP.


> Thank you.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> For additional commands, e-mail: users-help@httpd.apache.org
>
>

Hello,Thank you so much for your reply.Is the following sentence correct?

"If you did make mod_php thread safe, it wouldn’t change the fact that you’re going to blow up your memory usage as Apache will load PHP into every single worker process that spins up."



On Sun, Oct 1, 2023 at 12:26 PM, Deepak Goel<deicool@gmail.com> wrote:

On Sat, 30 Sept 2023, 12:26 Jason Long, <hack3rcon@yahoo.com.invalid> wrote:

Hello,
Is the following sentence correct?
"The way Apache loads PHP in its standard setup (with mod_php) compared to Nginx alone puts it at a disadvantage. You will see performance gains, particularly in memory usage, just by switching to Nginx, given you're using a PHP-driven application."


You will have to performance benchmark both Apache, Nginx to make a statement.
We use Apache with PHP.


Thank you.

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org

Hello,Thanks again.Why has Apache Foundation never tested Apache performance with Nginx?



On Sat, Sep 30, 2023 at 12:00 PM, Frank Gingras<thumbs@apache.org> wrote: There might be some online, however, due to the different architectures, they are not likely to be terribly useful. Do avoid the ones that bash needlessly either product.
On Sat, Sep 30, 2023 at 3:09?AM Jason Long <hack3rcon@yahoo.com.invalid> wrote:

Hello,
Thank you so much for your info.
Why are they trollish? I am curious to learn more.
Is there a fair comparison between Apache and Nginx?



On Saturday, September 30, 2023 at 10:35:12 AM GMT+3:30, Frank Gingras <thumbs@apache.org> wrote:





Additionally, your recent string of questions to this mailing list come off as a bit trollish.

On Sat, Sep 30, 2023 at 3:04?AM Frank Gingras <thumbs@apache.org> wrote:
> If any of the mod_php extensions are not thread-safe, you will need to use the prefork mpm, which will indeed bloat every httpd worker. This is not the ideal nor recommended configuration.
>
> Instead, use the event mpm and proxy_fcgi to pass the request to php-fpm.
>
> Alternatively, you can recompile php to be thread-safe and use event mpm with mod_php, which will give you a small execution speed advantage.
>
> The statement you posted is more or less FUD which leaves out very important details.
>
> On Sat, Sep 30, 2023 at 2:56?AM Jason Long <hack3rcon@yahoo.com.invalid> wrote:
>> Hello,
>> Is the following sentence correct?
>> "The way Apache loads PHP in its standard setup (with mod_php) compared to Nginx alone puts it at a disadvantage. You will see performance gains, particularly in memory usage, just by switching to Nginx, given you're using a PHP-driven application."
>>
>> Thank you.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
>> For additional commands, e-mail: users-help@httpd.apache.org
>>
>>
>


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org

I know that with nginx you can't configure your chain certificate separately, super annoying.

>
> Hello,
> Thanks again.
> Why has Apache Foundation never tested Apache performance with Nginx?
>
>
>
>
> On Sat, Sep 30, 2023 at 12:00 PM, Frank Gingras
> <thumbs@apache.org> wrote:
> There might be some online, however, due to the different
> architectures, they are not likely to be terribly useful. Do avoid the ones
> that bash needlessly either product.
>
> On Sat, Sep 30, 2023 at 3:09?AM Jason Long
> <hack3rcon@yahoo.com.invalid> wrote:
>
>
> Hello,
> Thank you so much for your info.
> Why are they trollish? I am curious to learn more.
> Is there a fair comparison between Apache and Nginx?
>
>
>
> On Saturday, September 30, 2023 at 10:35:12 AM GMT+3:30, Frank
> Gingras <thumbs@apache.org <mailto:thumbs@apache.org> > wrote:
>
>
>
>
>
> Additionally, your recent string of questions to this mailing
> list come off as a bit trollish.
>
> On Sat, Sep 30, 2023 at 3:04?AM Frank Gingras <thumbs@apache.org
> <mailto:thumbs@apache.org> > wrote:
> > If any of the mod_php extensions are not thread-safe, you will
> need to use the prefork mpm, which will indeed bloat every httpd worker.
> This is not the ideal nor recommended configuration.
> >
> > Instead, use the event mpm and proxy_fcgi to pass the request
> to php-fpm.
> >
> > Alternatively, you can recompile php to be thread-safe and use
> event mpm with mod_php, which will give you a small execution speed
> advantage.
> >
> > The statement you posted is more or less FUD which leaves out
> very important details.
> >
> > On Sat, Sep 30, 2023 at 2:56?AM Jason Long
> <hack3rcon@yahoo.com.invalid> wrote:
> >> Hello,
> >> Is the following sentence correct?
> >> "The way Apache loads PHP in its standard setup (with
> mod_php) compared to Nginx alone puts it at a disadvantage. You will see
> performance gains, particularly in memory usage, just by switching to
> Nginx, given you're using a PHP-driven application."
> >>
> >> Thank you.
> >>
> >> -------------------------------------------------------------
> --------
> >> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> <mailto:users-unsubscribe@httpd.apache.org>
> >> For additional commands, e-mail: users-help@httpd.apache.org
> <mailto:users-help@httpd.apache.org>
> >>
> >>
> >
>
>
> ----------------------------------------------------------------
> -----
> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> <mailto:users-unsubscribe@httpd.apache.org>
> For additional commands, e-mail: users-help@httpd.apache.org
> <mailto:users-help@httpd.apache.org>
>
>

B?KKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKCB??[??X???X?KK[XZ[?\?\??][??X???X?P?\X?K???B???Y][?[??[X[??K[XZ[?\?\??Z[?\X?K???B

On Wednesday 04 October 2023 at 20:48:19, Jason Long wrote:

> Hello,Thanks again.Why has Apache Foundation never tested Apache
> performance with Nginx?

I am not affiliated with the Apache Foundation in any way, but I would guess
that the primary reason is that one can make statistics say almost whatever
one wants them to, simply by selecting the data or analysis which supports the
desired outcome. Therefore nobody is going to trust numbers put out either by
the Apache Foundation, or by Nginx, showing how they compare against the
competition. I'm not saying that either of these organisations would be
lying, but they'd be expected to choose the tests and scenarios which show
them up in the most favourable comparative light possible.

A secondary reason is that one person's use of a web server is not the same as
another's, so any benchmarks showing Apache vs. Nginx would be idealistic and
almost certainly not what any specific real-world implementation would achieve.

Suppose you wanted to compare two makes of cars to find out which is "faster,
more secure and better" (to quote from the subject line of your email). Would
you want such a comparison to be done by manufacturer A, manufacturer B, or an
independent third party? No matter who it's done by, does their definition of
"better" match with yours (assuming you're a potential purchaser of one of the
cars)?


Antony.

--
The Free Software Foundation was formed on this day in 1985
https://www.fsf.org

Please reply to the list;
please *don't* CC me.

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org

On 2023-10-04 14:01, Antony Stone wrote:
> On Wednesday 04 October 2023 at 20:48:19, Jason Long wrote:
>
>> Hello,Thanks again.Why has Apache Foundation never tested Apache
>> performance with Nginx?
>
> I am not affiliated with the Apache Foundation in any way, but I would guess
> that the primary reason is that one can make statistics say almost whatever
> one wants them to, simply by selecting the data or analysis which supports the
> desired outcome. Therefore nobody is going to trust numbers put out either by
> the Apache Foundation, or by Nginx, showing how they compare against the
> competition. I'm not saying that either of these organisations would be
> lying, but they'd be expected to choose the tests and scenarios which show
> them up in the most favourable comparative light possible.
>
> A secondary reason is that one person's use of a web server is not the same as
> another's, so any benchmarks showing Apache vs. Nginx would be idealistic and
> almost certainly not what any specific real-world implementation would achieve.
>
> Suppose you wanted to compare two makes of cars to find out which is "faster,
> more secure and better" (to quote from the subject line of your email). Would
> you want such a comparison to be done by manufacturer A, manufacturer B, or an
> independent third party? No matter who it's done by, does their definition of
> "better" match with yours (assuming you're a potential purchaser of one of the
> cars)?

The more official, canonical reason is that NGINX is a commercial
company making an "open core" product, while the ASF is a non-profit.

The ASF cannot and does not want to compete with other products or
companies. It is not our mission, and we frankly do not care about
market shares or the likes. We are volunteers working on making a free
piece of software that can be used by whomever wants to use it.

>
>
> Antony.
>


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org

On 10/4/23 21:30, Daniel Gruno wrote:
> On 2023-10-04 14:01, Antony Stone wrote:
>> On Wednesday 04 October 2023 at 20:48:19, Jason Long wrote:
>>
>>> Hello,Thanks again.Why has Apache Foundation never tested Apache
>>> performance with Nginx?
>>
>> I am not affiliated with the Apache Foundation in any way, but I would
>> guess
>> that the primary reason is that one can make statistics say almost
>> whatever
>> one wants them to, simply by selecting the data or analysis which
>> supports the
>> desired outcome.  Therefore nobody is going to trust numbers put out
>> either by
>> the Apache Foundation, or by Nginx, showing how they compare against the
>> competition.  I'm not saying that either of these organisations would be
>> lying, but they'd be expected to choose the tests and scenarios which
>> show
>> them up in the most favourable comparative light possible.
>>
>> A secondary reason is that one person's use of a web server is not the
>> same as
>> another's, so any benchmarks showing Apache vs. Nginx would be
>> idealistic and
>> almost certainly not what any specific real-world implementation would
>> achieve.
>>
>> Suppose you wanted to compare two makes of cars to find out which is
>> "faster,
>> more secure and better" (to quote from the subject line of your
>> email).  Would
>> you want such a comparison to be done by manufacturer A, manufacturer
>> B, or an
>> independent third party?  No matter who it's done by, does their
>> definition of
>> "better" match with yours (assuming you're a potential purchaser of
>> one of the
>> cars)?
>
> The more official, canonical reason is that NGINX is a commercial
> company making an "open core" product, while the ASF is a non-profit.
>
> The ASF cannot and does not want to compete with other products or
> companies. It is not our mission, and we frankly do not care about
> market shares or the likes. We are volunteers working on making a free
> piece of software that can be used by whomever wants to use it.

Additionally comparing NGINX with Apache HTTPD server is comparing a
kitchen knife to a Swiss army knife... The choice depends on what you
need to do ;-)

>
>>
>>
>> Antony.
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
> For additional commands, e-mail: users-help@httpd.apache.org
>

--
Cheers

Jean-Frederic


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org