This sounds like syncronizing files in the file system, so rsync should
do the trick.
http://everythinglinux.org/rsync/
Google is your friend.
Scott
Mahabub Basha wrote:
> Hi
>
> I am the fan of Cherrypy. I am running CP 2.2.0 beta. I am running two
> cherrypy servers on two machines which is connected to LAN.
>
> one is master server and another one is backup server. Whatever
> changes updated in master should be updated in backup server also.
>
> For example I am storing the form data in dictionary. The dictionary
> should serialized as pickle.
>
> I have to push the pickle file to my backup server whatever changes
> made in form data.
>
> my master server running on 172.16.1.1 <http://172.16.1.1>
>
> my backup server running on 172.16.1.2 <http://172.16.1.2>
I think the idea is that other libraries used in conjunction with
CherryPy would be beneficial. For example, if you take a POST request
to your CherryPy app, you could load the data from it and send it to
another application via a HTTP request using something like httplib2.
While I am not sure of the CherryPy specifics, you can read from a
file like object in the WSGI environ dict using the 'wsgi.input' key
when you are trying to accept the POST'd data as a file (as opposed to
variables).
The CherryPy Essentials book discusses this in its discussion of
RESTful protocols. It is worth the $25 to have a PDF version (IMHO).
I realize this all might be a bit vague or unclear but hopefully, it
can give you a few ideas of terms to google :)
Eric
PS Often times the answer is very easy to find on google, yet the
terms are difficult to pin down. Good luck!
>
> - Basha.S
>
>
>
> >
>
Basically your application would open a connection to a queue server and
send a queue message which payload can be whatever. The queue server
will use either memcached or SQS to store the queue message.
Then other applications can read from the queue and fetch a queue
message to process it. A message can also be deleted of course.
The point is to avoid that your main application has to wait for your
other applications to process whatever data the main application generates.
Instead you decentralize the work in a nice and scalable way. You may
get inspiration :)
- Sylvain
[1] http://trac.defuze.org/wiki/bucker
[2] http://kamaelia.sourceforge.net/Home
Alberto José Rubert Escuder a écrit :
I think the better way is to create a conection (TCP or UDP, as you want but to send files i sugest UDP )...
-- Tim Roberts, ti...@probo.com Providenza & Boekelheide, Inc.