IMAP HILLWAY\ said:
>
> Now... this works fine for almost every URL I can think of apart from 2:
> http://www.acxiom.co.uk & http://www.acxiom.com
> These 2 both return "500 Internal Server Error" but work fine if you go
> to them with a browser!
bit of checking and it seems that they're using the
Nigel Hamilton wrote:
One advantage of the pipe is I can GZIP it as it goes ... server survival
is the first thing though time for a beer too! :-)
You're going to gzip the traffic through the pipes? Is that a
performance thing?
--
Andy Armstrong, Tagish
>
> Cool.. Dynamically adjusting number of clients based on load works well too,
> as you've said below.
> Best method for checking load on Linux, without incurring much system time, is
> to open /proc/loadavg and read that. If you seek to start of file and re-read(),
> you get the updated loadav
On Mon, Jun 02, 2003 at 11:59:22AM +0100, Alex Hudson wrote:
> On Mon, Jun 02, 2003 at 05:48:22AM -0500, Nigel Hamilton wrote:
> > > I suspect i get called heretical for this, but I actually prefer to use
> > > threads when it comes to these things.
> > > a) better memory efficiency
> > > b) easie
On Mon, Jun 02, 2003 at 05:48:22AM -0500, Nigel Hamilton wrote:
> > In what way are you limiting the total number of processes?
> >
>
> I?ve just implemented a pre-forking server so it is limited to the number
> of pre-forked servers * 10 forks per server - this has improved things a
> bit.
Co
On Mon, 2 Jun 2003, Dave Hodgkinson wrote:
> On Mon, 2003-06-02 at 10:42, the hatter wrote:
> > Anyone else going to be there ? Any existing plans for moungers to gather
> > in small corners and discuss current industry trends in the key sectors,
> > buffy and beer ?
>
> I'll be there later on, p
On Mon, 2003-06-02 at 10:42, the hatter wrote:
> Anyone else going to be there ? Any existing plans for moungers to gather
> in small corners and discuss current industry trends in the key sectors,
> buffy and beer ?
I'll be there later on, prolly Thursday. Save me some l00t.
--
Dave Hodgkinson
On 29 May 2003 at 6:59, Toby Corkindale wrote:
> On Tue, May 27, 2003 at 09:59:32PM +0100, David Cantrell wrote:
> > I too am an advocate of old-style spelling. Unfortunately, this keyboard
> > lacks eth, thorn, etc.
>
> vim probably allows access via the Ctrl-K thingies?
TH and th, D- and d-
On Sun, Jun 01, 2003 at 09:18:26PM +0100, Lusercop wrote:
> > $SIG{CHLD} = 'IGNORE';
>
> How is that even remotely the same? In the first case, you end up with no
> zombies, in the second case you end up with a load of them? Admittedly, if
[snip]
>
> Toby, I'd have expected better from you. :-)
>
On Mon, Jun 02, 2003 at 05:48:22AM -0500, Nigel Hamilton wrote:
> > I suspect i get called heretical for this, but I actually prefer to use
> > threads when it comes to these things.
> > a) better memory efficiency
> > b) easier and more advanced control over queueing and message handling
> > c) a
>
> In what way are you limiting the total number of processes?
>
I´ve just implemented a pre-forking server so it is limited to the number
of pre-forked servers * 10 forks per server - this has improved things a
bit.
> You probably know anyway, but remember that 20 processes that operate wit
On Sun, Jun 01, 2003 at 05:41:38PM -0500, Nigel Hamilton wrote:
> Thanks to everyone who has sent suggestions ... it's tricky ... it does
> appear that the forking goes out of control... despite the signal handling
> code even if I get this to work robustly I?m still going to look at
> other a
On Mon, Jun 02, 2003 at 09:42:57AM +, the hatter wrote:
> Anyone else going to be there ? Any existing plans for moungers to gather
> in small corners and discuss current industry trends in the key sectors,
> buffy and beer ?
I am hiding from anything involving buffy, for fear of spoilers. I'
Anyone else going to be there ? Any existing plans for moungers to gather
in small corners and discuss current industry trends in the key sectors,
buffy and beer ?
the hatter
On Mon, 2003-06-02 at 01:04, Lusercop wrote:
> > That's what I mean by 'special flag', although I can't see where it gets
> > set. People mention it interferes with sigaction(), but little about
> > what actually happens. I looked in signal.c and fork.c and it's not
> > obvious, unless it's somethi
On Sun, Jun 01, 2003 at 10:34:19PM +0100, Alex Hudson wrote:
> On Sun, Jun 01, 2003 at 10:09:00PM +0100, Lusercop wrote:
> > > Given the process ids are sequential in my little test program, it's not
> > > Perl doing some daemonlike double-fork(), so I guess some special flag
> > > gets set somewhe
On Sun, 1 Jun 2003, Lusercop wrote:
>
> But this seems a reasonable suggestion. I've never quite figured out why the
> signal handlers in perl are interruptible. This seems to be a bug, but I'm
> not quite sure. Signals have never been my strong point, I normally know
> enough to get by...
>
ISTR
Thanks to everyone who has sent suggestions ... it's tricky ... it does
appear that the forking goes out of control... despite the signal handling
code even if I get this to work robustly I´m still going to look at
other alternatives - under load the response time is too variable.
Thanks aga
On Sun, Jun 01, 2003 at 10:09:00PM +0100, Lusercop wrote:
> > Given the process ids are sequential in my little test program, it's not
> > Perl doing some daemonlike double-fork(), so I guess some special flag
> > gets set somewhere or something.. dunno ;) I didn't think this
> > behaviour existed
On Sun, Jun 01, 2003 at 09:43:13PM +0100, Alex Hudson wrote:
> On Sun, Jun 01, 2003 at 09:18:26PM +0100, Lusercop wrote:
> > How is that even remotely the same? In the first case, you end up with no
> > zombies, in the second case you end up with a load of them?
> Hmm, unless I'm missing something,
On Sun, Jun 01, 2003 at 09:18:26PM +0100, Lusercop wrote:
> > > $SIG{CHLD} = sub { while (waitpid(-1,WNOHANG) > 0 ) { } };
> > If that's all you are doing, then is there anything wrong with
> > changing it to this?
> > $SIG{CHLD} = 'IGNORE';
>
> How is that even remotely the same? In the first c
On Sun, Jun 01, 2003 at 09:29:08PM +0100, Dirk Koopman wrote:
> while (in some idle loop or other) {
> if ($death) {
> $death-- while waitpid(-1, WNOHANG);
> }
> }
I'd be inclined to swap the if and while:
| while (in some idle loop or other && $death) {
| $dea
On Sun, 2003-06-01 at 17:53, Alex Hudson wrote:
> On Sun, Jun 01, 2003 at 07:57:11AM -0700, Toby Corkindale wrote:
> > > sub reap {
> > > while (waitpid(-1,WNOHANG);
> > > $SIG{CHLD} = \&reap;
> > > }
> > > $SIG{CHLD} = \&reap;
> >
> > Depends on operating system. I think Linux and some other
On Sun, Jun 01, 2003 at 03:53:30PM +0100, Alex McLintock wrote:
> Hi folks,
> I'm still not sorted out for colocation of my mod_perl based sites (and
> java too)
http://www.blackcatnetworks.co.uk/ (do you mean co-location, or do you mean
managed hosting?)
--
Lusercop.net - LARTing Lusers everyw
On Sun, Jun 01, 2003 at 07:41:45AM -0700, Toby Corkindale wrote:
> On Sun, Jun 01, 2003 at 09:15:27AM -0500, Nigel Hamilton wrote:
> > Here is the code that is trying to reap child processes:
> > use POSIX 'WNOHANG';# see Stein's idiom p. 305 Network Programming with Perl
> > # kill all zombie'
On Sunday, June 1, 2003 19:44 +0100 Peter Sergeant <[EMAIL PROTECTED]>
wrote:
I can highly recommend http://www.bytemark-hosting.co.uk/ - not
colocation, but virtual linux machines. Service has been absolutely
first class
Pity they don't answer their mail.
--
David Cantrell
Beekeeping is like
> > Depends on operating system. I think Linux and some other modern systems
> > maintain the setting, but others (UnixWare?) require it to be re-set.
>
> Hmm, you're right - force of habit, clearly :/ SCO OpenSewer r5 doesn't need
> it, don't know about UnixWare. Out of all of the, I would think
> I've been trying to find a cheap alternative to co-location, but they all
> seem to be a right PITA.
I can highly recommend http://www.bytemark-hosting.co.uk/ - not
colocation, but virtual linux machines. Service has been absolutely
first class
+Pete
On Sun, Jun 01, 2003 at 07:57:11AM -0700, Toby Corkindale wrote:
> > sub reap {
> > while (waitpid(-1,WNOHANG);
> > $SIG{CHLD} = \&reap;
> > }
> > $SIG{CHLD} = \&reap;
>
> Depends on operating system. I think Linux and some other modern systems
> maintain the setting, but others (UnixWare
Have you seen http://www.memset.com/ (or .co.uk)
I've not used them, but tempted. They aren't a co-lo, but given the prices,
maybe you could simply get another box and save the taxi fare? They host
smoothwall.org, which may or may not be a recommendation.
> Can anyone suggest a way that I can c
Toby Corkindale sent the following bits through the ether:
> Is the front-end important to be written in TT, or rather the publish end?
The front end is fine (and very pretty) as it is. However, on the
publish end it'd be nice if there was an easy-to-use TT burner.
Of course, I mean that in the
On Sun, Jun 01, 2003 at 03:36:38PM +0100, Alex Hudson wrote:
> > Any ideas on how to reap subprocesses robustly? ... If I can get
> > the processes to be reaped correctly the machine will be able to stay
> > up ...
>
> Don't you have to reinstate the handler once it has been used? E.g.;
>
>
On Fri, May 30, 2003 at 07:26:52PM +0100, Nicholas Clark wrote:
> > Anyway, just noticed that there is a Apache::Template module already, which
> > appears to do much of what you were after. I think?
>
> So can they be twisted into writing a TT front end for Bricolage?
Is the front-end important
Hi folks,
I'm still not sorted out for colocation of my mod_perl based sites (and
java too)
I've been trying to find a cheap alternative to co-location, but they all
seem to be a right PITA.
(I think I would still like an ourshack.com acount if someone will
recommend me :-)
So I'm looking at
On Sun, Jun 01, 2003 at 09:15:27AM -0500, Nigel Hamilton wrote:
> Here is the code that is trying to reap child processes:
>
> use POSIX 'WNOHANG';# see Stein's idiom p. 305 Network Programming with Perl
>
> # kill all zombie's including hanging zombies, needed for cleaning up
> # rogue pr
On Sun, Jun 01, 2003 at 09:15:27AM -0500, Nigel Hamilton wrote:
> # kill all zombie's including hanging zombies, needed for cleaning up
> # rogue processes
> $SIG{CHLD} = sub { while (waitpid(-1,WNOHANG) > 0 ) { } };
>
> Any ideas on how to reap subprocesses robustly? ... If I can get
> t
On Sun, 1 Jun 2003, Simon Wistow wrote:
> On Sun, Jun 01, 2003 at 08:03:52AM -0500, Nigel Hamilton said:
> > Hi,
> > I´ve got a mini-weberver
>
> And it's been Register-ed? (that doesn't sound nearly as natural as
> Slashdotted)
>
> Nice write up in El Reg. Any ETA on the site coming back u
On Sat, May 31, 2003 at 11:45:43PM +0100, David M. Wilson wrote:
> On Fri, May 30, 2003 at 09:19:09AM -0700, Toby Corkindale wrote:
>
> > Recompiling wget with --without-ssl fixes the problem. But it still sounds
> > like a bug.
>
> Without checking, are you getting the "-m" and "-r" options conf
38 matches
Mail list logo
