Hi Theoretically, if you replace headers with Mutivalue map, the areas which gonna impact are TargetResponse and SourceResponses (basically all the places where the headers getting consumed)
FYI Ravi: Before me or Kasun apply this to synapse,* please make sure you have run all the Automation tests and integration tests, *make sure this modification does not introduced any new anomalies, so take your time and run AM and integration tests..and provide us the surefire test reports. * "**I'm not sure what this may have broken. "* hope you got the answer cheers Dushan On Tue, Mar 12, 2013 at 9:45 AM, Nirmal Fernando <[email protected]> wrote: > Hi Ravi, > > > On Mon, Mar 11, 2013 at 8:58 AM, Ravi Undupitiya <[email protected]> wrote: > >> Hello, >> >> Adding ESBJAVA-1659 to passthru did the trick - took me a while to >> understand how the NIO patch worked and also how to add it to passthru. >> > > Nice to hear that fix worked. (I didn't get a chance to look at your > patch, but I know that the NIO patch, I created, is working :-)) > >> >> I've attached the patch here. This should be applied to Synapse >> 2.1.1-wso2v2. >> > > I think it's better if you could merge this to trunk. Synapse 2.1.1-wso2v2 > is already released and you don't need to patch, since the bug we're > solving is of S2 and we gonna cut S2 from trunk. > >> >> I'm not sure what this may have broken. >> > > Well, get the help from Dushan, Kasun et al. and do a round of test, > please :-) > > >> >> >> Thanks, >> Ravi >> >> On Thu, Mar 7, 2013 at 6:06 AM, Nirmal Fernando <[email protected]> wrote: >> >>> >>> On Mar 6, 2013 9:15 PM, "Ravi Undupitiya" <[email protected]> wrote: >>> > >>> > Hello, >>> > >>> > >>> > >>> > When the Stratos ELB is proxying third party applications - some of >>> these applications send several Set-Cookie headers in one response. >>> > >>> > RFC2109 states: >>> > >>> > An origin server may include multiple Set-Cookie headers in a >>> > response. Note that an intervening gateway could fold multiple such >>> > headers into a single header. >>> > >>> > >>> > Which means we should be able to fold multiple cookies into one >>> Set-Cookie. However this specification is obsolete and the more >>> up-to-date version (RFC6265) states: >>> > >>> > >>> > Origin servers SHOULD NOT fold multiple Set-Cookie header fields into >>> > a single header field. The usual mechanism for folding HTTP headers >>> > fields (i.e., as defined in [RFC2616]) might change the semantics of >>> > the Set-Cookie header field because the %x2C (",") character is used >>> > by Set-Cookie in a way that conflicts with such folding. >>> > >>> > >>> > >>> > I suspect this to be the issue when proxying Wordpress via the ELB - >>> Multiple Set-Cookies are ignored by the current Synapse implementation. >>> > >>> > I tried, making the synapse implementation fold multiple cookies (as >>> per the older RFC2109) using "," but it didn't do anything different. >>> > >>> > >>> > Did the wordpress login work in previous versions? >>> > >>> > Also I have been implementing >>> https://wso2.org/jira/browse/ESBJAVA-1659 for the passthru transport - >>> but if I understand correctly this only solves the issue of a header having >>> multiple values and not multiple headers by the same name? >>> >>> If you read the test case there, you'll see that this is the exact issue. >>> > >>> > >>> > Thanks, >>> > >>> > Ravi >>> > >>> > >>> > >>> > >>> > >>> >> >> > > > -- > > Thanks & regards, > Nirmal > > Software Engineer- Platform Technologies Team, WSO2 Inc. > Mobile: +94715779733 > Blog: http://nirmalfdo.blogspot.com/ > > <http://nirmalfdo.blogspot.com/> > -- Dushan Abeyruwan *Senior Software Engineer* *Integration Technologies Team* *Member of Integration MC* *WSO2 Inc. http://wso2.com/* *Mobile:(+94)714408632*
_______________________________________________ Dev mailing list [email protected] http://wso2.org/cgi-bin/mailman/listinfo/dev
