In memory persistence - used as CGI
Alex Martelli
aleaxit at yahoo.com
Tue Oct 31 06:48:33 EST 2000
More information about the Python-list mailing list
Tue Oct 31 06:48:33 EST 2000
- Previous message (by thread): In memory persistence - used as CGI
- Next message (by thread): a simple example of Stackless Python
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
"Robin Dunn" <robin at alldunn.com> wrote in message news:8tl1i6$fuv$1 at nnrp1.deja.com... > In article <8tkn1b09kg at news1.newsguy.com>, > "Alex Martelli" <aleaxit at yahoo.com> wrote: > > <youngcho_sd at my-deja.com> wrote in message > > news:8tk9aq$p80$1 at nnrp1.deja.com... > > > I'm using Python as a CGI language (for use both [snip] > > No, each CGI 'hit' is served by a completely separate > > process (I think this holds for both IIS and Apache), [snip] > > have to go to disk, and/or use cookies. > > Or use something like FastCGI, but then you have to funnel all requests > through a single process which is fine for light or medium loads, but > can cause throughput issues at a high load. If CGI is not required, then multiple processes might still be acceptable -- the mmap builtin module (new in 2.0, I believe) will allow them to share the memory areas to which objects get 'in-memory persisted'. It works on both Unix and Windows. The memory may be backed by the filesystem rather than the pagingfile, but that shouldn't be an issue. Alex
- Previous message (by thread): In memory persistence - used as CGI
- Next message (by thread): a simple example of Stackless Python
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Python-list mailing list