Export spaces vua command line(using SOAP remote API) times out

Maya Srinivasan

Hi !

I am working with a standalone Confluence installations (3.5.6) and I'd like to export spaces from the commandline (Confluence CLI seems to be the answer).

When I try to export a reasonably large wiki space using Confluence CLI (which uses SOAP Remote API), I get:

Remote error: ; nested exception is:
Java.net.SocketTimeoutException: Read timed out

It times out in 10mins. Exactly.

Export space does work from within Confluence, and takes about 22 minutes. I've tested the export with a smaller space, and it works just fine through the CLI.

What is the right way to enable an end user to change the default timeout for all TCP connections or those initiated in a JVM so that these longer running tasks can finish and return? And, more importantly, is this the way to do it?

Has anyone had success exporting large spaces via command line?

Thanks.Maya

Adam Laskowski

Maya,

I don't quite know how to get around this issue with the CLI tool for Confluence, but I was able to overcome this issue with a Python script that might be useful:

#!/usr/bin/python

import sys, string, xmlrpclib, re, httplib, time

#http timeout override
class TimeoutHTTPConnection( httplib.HTTPConnection ):
   def connect( self ):
       httplib.HTTPConnection.connect( self )
       self.sock.settimeout( self.timeout )

class TimeoutHTTP( httplib.HTTP ):
   _connection_class = TimeoutHTTPConnection
   def set_timeout( self, timeout ):
       self._conn.timeout = timeout

class TimeoutTransport( xmlrpclib.Transport ):
    def __init__( self, timeout = 10, *l, **kw ):
        xmlrpclib.Transport.__init__( self, *l, **kw )
        self.timeout = timeout
    def make_connection( self, host ):
        conn = TimeoutHTTP( host )
        conn.set_timeout( self.timeout )
        return conn

class TimeoutServerProxy( xmlrpclib.ServerProxy ):
    def __init__( self, uri, timeout = 10, *l, **kw ):
        kw['transport'] = TimeoutTransport( timeout = timeout, use_datetime = kw.get( 'use_datetime', 0 ) )
        xmlrpclib.ServerProxy.__init__( self, uri, *l, **kw )


def runExport(newtimeout):
	#server = xmlrpclib.ServerProxy('http://localhost:8080/rpc/xmlrpc')

	try:
		print ("Starting at: "+time.asctime())
		server = TimeoutServerProxy( 'http://localhost:8080/rpc/xmlrpc', timeout = newtimeout )
		token = server.confluence1.login( 'admin', 'admin' )
		key = raw_input( "Enter space key: " )
		vartype = raw_input( "Eport type (xml or html): " )
		if "xml" in vartype.lower():
			type = 'TYPE_XML'
		else: 
			if "html" in vartype.lower():
				type = 'TYPE_HTML'
			else: print ("Invalid output type.")
		string = server.confluence1.exportSpace( token, key, type )
		exit( string )
	except Exception as e:
		print (time.asctime())
		print (e)
	finally:
		print (time.asctime())
	
toset = int(raw_input( "Set Timeout Value in Minutes: " ))*60
runExport(toset)

You will need to change the details in 'server' and 'token' to match your server URL and login details. When you run the script, set a timeout value in minutes and specify the space key and xml and it should eventually spit out a URL. In local testing, a space with 35000 unique pages, each with one image attached, finished in 16 minutes. Mileage will vary, but you can get a feel for how long you can expect it to take by running a space export in the Confluence UI to establish your baseline.