On Monday, Feb 24, 2003, at 19:47 Europe/London, Jim Morrison
[Mailing-Lists] wrote:
What I thought was that if I kept the handle to the parsed XML open
somewhere else then I would be able to use it.. so a separate process
does the parsing and keeps hold of the handles of the currently
'shared'
>
> Jim Morrison [Mailing-Lists] wrote:
> > Sniff.. I've kind of got something working... Enough such that one
> > httpd can request an XYZ, and if a second httpd comes along a little
> > later and requests the same XYZ then it will get it from shared
memory.
>
> I hope you used one of the modul
>> Hmm.. Yes, it sounds pretty sketchy to me too! Immediately what I am
>> playing with is the idea of keeping parsed XML (XML::LibXML)in memory
>> between requests. Is this a completely barmy idea?
> Probably, because you'll confuse XML::LibXML's garbage collector.
Sniff.. I've kind of got s
On Mon, 2003-02-24 at 05:40, Jim Morrison [Mailing-Lists] wrote:
> Hmm.. Yes, it sounds pretty sketchy to me too! Immediately what I am
> playing with is the idea of keeping parsed XML (XML::LibXML)in memory
> between requests. Is this a completely barmy idea?
Caching is a good idea, but it's no
Perrin,
> I suggest you look at IPC::MM or IPC::Shareable. IPC::Shareable is
more
> transparent, but IPC::MM has better performance. IPC::MM simply
creates
> a hash in shared memory and lets you write to it. Either of these
will
> allow you to share data between processes.
Thanks, I'll have a
On Sun, 2003-02-23 at 10:46, Jim Morrison [Mailing-Lists] wrote:
> Having spent the w/e getting to grips with startup.pl's and the such I'm
> beginning to discover that it's only possible to share read-only memory
> and as soon as you write to memory it splits off.. ..
The situation with forked A
