SimpleHTTPReque stHandler handling long lasting requests problem

[Andy Leszczynski]

I need a HTTP server handling long lasting requests e.g. 10-30 seconds.
Below is a pice of the code. In order to make the server reponsive while
handling othere requests I use SocketServer.Th readingMixIn.

However the problem is the it does not work out. I checked thet a new
thread is created for each new connection new, but the main loop seems
to be frozen until the prevoius handling ends.

What could go wrong?

Thanks, Andy

* * *

import os
import time
import BaseHTTPServer
import SocketServer
import threading
import sys

class
SimpleHTTPReque stHandler(BaseH TTPServer.BaseH TTPRequestHandl er):
def do_GET(self):
response="aaaaa aaaaaaaaaaa"+st r(time.time())
self.send_respo nse(200)
self.send_heade r("Content-type",'text/plain')

self.send_heade r("Content-Length",len(res ponse))
self.end_header s()
time.sleep(10) #simulation of the processing
self.wfile.writ e(response)

def do_POST(self):
self.do_GET()

class
myWebServer(Soc ketServer.Threa dingMixIn,BaseH TTPServer.HTTPS erver):
pass

if __name__ == '__main__':
server_address = ('',80)
httpd=myWebServ er(server_addre ss,SimpleHTTPRe questHandler)
sa=httpd.socket .getsockname()
print "Serving HTTP on", sa[0], "port", sa[1],"..."
httpd.serve_for ever()

[Andy Leszczynski]

Sorry for questioning Python :-) - it turned out that this is a problem
with Mozilla. For some reason it holds up with opening second connection
to given host until the previous one is completed. Interestingly enough,
IE works better with Python multi threaded server in that regard.

Thx, A.

[Steve Holden]

Andy Leszczynski wrote:

Sorry for questioning Python :-) - it turned out that this is a problem
with Mozilla. For some reason it holds up with opening second connection
to given host until the previous one is completed. Interestingly enough,
IE works better with Python multi threaded server in that regard.

Thx, A.

Try switching keepalives off, or falling back to HTTP 1.0 - ironically
it may be the attempt to use the same connection for both pieces of
content that holds things up.

regards
Steve
--
Steve Holden [phone] [phone]
Holden Web LLC http://www.holdenweb.com/
Python Web Programming http://pydish.holdenweb.com/

[Andy Leszczynski]

Steve Holden wrote:

Andy Leszczynski wrote:

Sorry for questioning Python :-) - it turned out that this is a
problem with Mozilla. For some reason it holds up with opening second
connection to given host until the previous one is completed.
Interestingly enough, IE works better with Python multi threaded
server in that regard.

Thx, A.

Try switching keepalives off, or falling back to HTTP 1.0 - ironically
it may be the attempt to use the same connection for both pieces of
content that holds things up.

regards
Steve

I tested it before and it did not work either. Have to try HTTP 1.0 thouh.

Thanks, A.

[Andy Leszczynski]

Steve Holden wrote:

Andy Leszczynski wrote:

Sorry for questioning Python :-) - it turned out that this is a
problem with Mozilla. For some reason it holds up with opening second
connection to given host until the previous one is completed.
Interestingly enough, IE works better with Python multi threaded
server in that regard.

Thx, A.

Try switching keepalives off, or falling back to HTTP 1.0 - ironically
it may be the attempt to use the same connection for both pieces of
content that holds things up.

regards
Steve

I tested it before and it did not work either. Have to try HTTP 1.0 thouh.

Thanks, A.