Mailing List Archive: Re: [Dbmail] memory problem on RedHat 9 and dbmail 1.1

Re: [Dbmail] memory problem on RedHat 9 and dbmail 1.1

Oct 24, 2003, 12:20 AM

Post #1 of 5 (917 views)

<we should move this thread to dbmail-dev>

I for one would vote against having those patches be accepted into CVS,
and I imagine Ilja wont commit them as they are.

Don't get me wrong though; inetd functionality is a valuable attribute,
but I don't think they should be acquired by forking existing files.

Personally, since I'm working on pre-forking functionality ala apache,
I'm pretty certain server.c is where such changes should occur.

And we should definitely allow admins to select such settings runtime:
SERVERTYPE=<inetd|standalone>

Matthew T. O'Connor wrote:
> On Thu, 2003-10-23 at 17:25, Matt Dickinson wrote:
>
>>Matthew T. O'Connor wrote:
>>
>>>I assume these patches are all against CVS (aka 2.0 branch).
>>
>>No, sorry. 1.2 branch.
>>I'll take a look @ 2.0 branch tomorrow.
>
>
> Good as, I would assume that these changes would not be accepted into
> CVS for 1.2, but might (I think should) be accepted for 2.0.
>
> _______________________________________________
> Dbmail mailing list
> Dbmail@dbmail.org
> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>

--
________________________________________________________________
Paul Stevens mailto:paul@nfg.nl
NET FACILITIES GROUP PGP: finger paul@nfg.nl
The Netherlands________________________________http://www.nfg.nl

Re: Re: [Dbmail] memory problem on RedHat 9 and dbmail 1.1 [ In reply to ]

ilja at ic-s

Oct 24, 2003, 12:25 AM

Post #2 of 5 (883 views)

Permalink

Hi,

my thoughts exactly :)

Any luck with the pre-forking stuff yet?

cheers,
Ilja

On Friday, Oct 24, 2003, at 09:20 Europe/Amsterdam, Paul J Stevens
wrote:

> <we should move this thread to dbmail-dev>
>
> I for one would vote against having those patches be accepted into
> CVS, and I imagine Ilja wont commit them as they are.
>
> Don't get me wrong though; inetd functionality is a valuable
> attribute, but I don't think they should be acquired by forking
> existing files.
>
> Personally, since I'm working on pre-forking functionality ala apache,
> I'm pretty certain server.c is where such changes should occur.
>
> And we should definitely allow admins to select such settings runtime:
> SERVERTYPE=<inetd|standalone>
>
>
> Matthew T. O'Connor wrote:
>> On Thu, 2003-10-23 at 17:25, Matt Dickinson wrote:
>>> Matthew T. O'Connor wrote:
>>>
>>>> I assume these patches are all against CVS (aka 2.0 branch).
>>>
>>> No, sorry. 1.2 branch. I'll take a look @ 2.0 branch tomorrow.
>> Good as, I would assume that these changes would not be accepted into
>> CVS for 1.2, but might (I think should) be accepted for 2.0.
>> _______________________________________________
>> Dbmail mailing list
>> Dbmail@dbmail.org
>> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>
> --
> ________________________________________________________________
> Paul Stevens mailto:paul@nfg.nl
> NET FACILITIES GROUP PGP: finger paul@nfg.nl
> The Netherlands________________________________http://www.nfg.nl
>
> _______________________________________________
> Dbmail-dev mailing list
> Dbmail-dev@dbmail.org
> http://twister.fastxs.net/mailman/listinfo/dbmail-dev
>
--
IC&S
Koningsweg 4
3582 GE UTRECHT

PGP-key:
http://www.ic-s.nl/keys/ilja.txt

Re: Re: [Dbmail] memory problem on RedHat 9 and dbmail 1.1 [ In reply to ]

paul at nfg

Oct 24, 2003, 4:13 AM

Post #3 of 5 (891 views)

Permalink

Ilja Booij wrote:
> Hi,
>
> my thoughts exactly :)
>
> Any luck with the pre-forking stuff yet?

I'm trying to setup a global scoreboard where I can track state for all
children. I'm using shared memory to do this. I've studied the apache
code, but I want to steer clear of thread for now. Stevens's Advanced
Programming has been on my desk for the last couple of days now, and I'm
making steady progress.

>
> cheers,
> Ilja
>
> On Friday, Oct 24, 2003, at 09:20 Europe/Amsterdam, Paul J Stevens wrote:
>
>> <we should move this thread to dbmail-dev>
>>
>> I for one would vote against having those patches be accepted into
>> CVS, and I imagine Ilja wont commit them as they are.
>>
>> Don't get me wrong though; inetd functionality is a valuable
>> attribute, but I don't think they should be acquired by forking
>> existing files.
>>
>> Personally, since I'm working on pre-forking functionality ala apache,
>> I'm pretty certain server.c is where such changes should occur.
>>
>> And we should definitely allow admins to select such settings runtime:
>> SERVERTYPE=<inetd|standalone>
>>
>>
>> Matthew T. O'Connor wrote:
>>
>>> On Thu, 2003-10-23 at 17:25, Matt Dickinson wrote:
>>>
>>>> Matthew T. O'Connor wrote:
>>>>
>>>>> I assume these patches are all against CVS (aka 2.0 branch).
>>>>
>>>>
>>>> No, sorry. 1.2 branch. I'll take a look @ 2.0 branch tomorrow.
>>>
>>> Good as, I would assume that these changes would not be accepted into
>>> CVS for 1.2, but might (I think should) be accepted for 2.0.
>>> _______________________________________________
>>> Dbmail mailing list
>>> Dbmail@dbmail.org
>>> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>>
>>
>> --
>> ________________________________________________________________
>> Paul Stevens mailto:paul@nfg.nl
>> NET FACILITIES GROUP PGP: finger paul@nfg.nl
>> The Netherlands________________________________http://www.nfg.nl
>>
>> _______________________________________________
>> Dbmail-dev mailing list
>> Dbmail-dev@dbmail.org
>> http://twister.fastxs.net/mailman/listinfo/dbmail-dev
>>
> --
> IC&S
> Koningsweg 4
> 3582 GE UTRECHT
>
> PGP-key:
> http://www.ic-s.nl/keys/ilja.txt
>
>
> _______________________________________________
> Dbmail-dev mailing list
> Dbmail-dev@dbmail.org
> http://twister.fastxs.net/mailman/listinfo/dbmail-dev
>

--
________________________________________________________________
Paul Stevens mailto:paul@nfg.nl
NET FACILITIES GROUP PGP: finger paul@nfg.nl
The Netherlands________________________________http://www.nfg.nl

Re: [Dbmail] memory problem on RedHat 9 and dbmail 1.1 [ In reply to ]

matthew at zeut

Oct 24, 2003, 7:06 AM

Post #4 of 5 (891 views)

Permalink

On Fri, 2003-10-24 at 03:20, Paul J Stevens wrote:
> I for one would vote against having those patches be accepted into CVS,
> and I imagine Ilja wont commit them as they are.
>
> Don't get me wrong though; inetd functionality is a valuable attribute,
> but I don't think they should be acquired by forking existing files.
>
> Personally, since I'm working on pre-forking functionality ala apache,
> I'm pretty certain server.c is where such changes should occur.
>
> And we should definitely allow admins to select such settings runtime:
> SERVERTYPE=<inetd|standalone>

All I was saying is that I think some type of xinetd functionality
should be added to the 2.0 branch. If this implmentation (patch)
doesn't cut the mustard, then fine we need to rework it, but I would
like to see (x)inetd supported, somehow.

Matthew

RE: Re: [Dbmail] memory problem on RedHat 9 and dbmail 1.1 [ In reply to ]

lists at alpha345

Oct 24, 2003, 9:28 AM

Post #5 of 5 (894 views)

Permalink

Matthew T. O'Connor wrote:
> On Fri, 2003-10-24 at 03:20, Paul J Stevens wrote:
>> I for one would vote against having those patches be accepted into
>> CVS, and I imagine Ilja wont commit them as they are.
>>
>> Don't get me wrong though; inetd functionality is a valuable
>> attribute, but I don't think they should be acquired by forking
>> existing files.
>>
>> Personally, since I'm working on pre-forking functionality ala
>> apache, I'm pretty certain server.c is where such changes should
>> occur.
>>
>> And we should definitely allow admins to select such settings
>> runtime: SERVERTYPE=<inetd|standalone>
>
> All I was saying is that I think some type of xinetd functionality
> should be added to the 2.0 branch. If this implmentation (patch)
> doesn't cut the mustard, then fine we need to rework it, but I would
> like to see (x)inetd supported, somehow.
>

I agree with both of you.. I never intended the things I wrote to be more
than a proof of concept, and I didn't implement into the existing structure
in order that I minimised stuff that I might have broken.

Since there is a demand (other than just me) for (x)inetd support, I'll have
a look at the 2.0 tree and see where I can make a start at a sensible patch
for that, and hopefully get it incorporated into the tree.

Matt

PS: sorry Moderator, I seem to send stuff from the wrong account all too
frequently :-(