Differences between revisions 4 and 5
Revision 4 as of 2007-06-13 21:27:34
Size: 18644
Editor: Deepal
Revision 5 as of 2009-09-20 22:48:48
Size: 18644
Editor: localhost
Comment: converted to 1.6 markup
No differences found!
  • [09:00] chathura has joined

[09:00] azeez_ has disconnected: "Leaving"

  • [09:01] srinathp has joined [09:01] azeez has joined [09:06] gdaniels has joined [09:09] Deepal: so what is the paln [09:09] Deepal: plan today [09:09] Deepal: going through JIRAs ..... ? [09:09] gdaniels: JIRA JIRA [09:09] Deepal: or shall we try to do some scanerios and see whethere we are getting the [09:10] Deepal: - correct responses [09:10] Deepal: - Correct faults [09:10] gdaniels: what kind of scenarios [09:10] gdaniels: ah

    [09:10] gdaniels: sounds like a test-case writing session to me :) [09:10] srinathp: we are yet to go though the server side [09:11] Deepal: I really need to make sure we do handle fault corretly

    [09:11] Deepal: glen : yes it is :) [09:11] gdaniels: Plus we need to carve out some time in the next few days to attack our fault messages [09:11] gdaniels: (use getMessage(), make the strings more informative, etc) [09:11] Deepal: in the server side we can thw exceptions in deferent places and see whether client get them [09:16] Deepal: hmm , seems like we are going to start with JIRA

    [09:16] Deepal: I will be working on https://issues.apache.org/jira/browse/AXIS2-2744

    [09:17] Deepal: I am almost done with that :) [09:18] saminda has joined [09:19] azeez: hi guys [09:19] azeez: Deepal, I have a question for u [09:20] Deepal: fire [09:20] azeez: u replied to my mail saying that the Abstract MR can call the replicateState method after invoking biz logic [09:20] Deepal: yes [09:20] azeez: but there is no invoke biz logic call from the abstract MR

    [09:21] Deepal: if you look at AbstractInMessageReceiver [09:21] azeez: yes.. [09:21] Deepal: you will understand what I am talking abt [09:21] gdaniels just replied to that thread also [09:21] azeez: those places I have invoked [09:22] azeez: all of our MRs I have done exactly what u have mentioned [09:23] Deepal: where do you want to call that method [09:23] Deepal: is it b4 calling the service or after calling the service [09:23] azeez: after calling the service

    [09:23] Deepal: ok then have a look at AbstractInOutSyncMessageReceiver [09:23] Deepal: there you can call that method b4 you call engine.send [09:24] Deepal: I know if we are going to do so we have a number of places to change [09:24] azeez: try { [09:24] azeez: invokeBusinessLogic(msgContext, outMsgContext); [09:24] azeez: replicateState(msgContext); [09:24] azeez: } finally { [09:24] azeez: restoreThreadContext(tc); [09:24] azeez: } [09:24] azeez: not many places deepal [09:24] azeez: few classes [09:25] Deepal: hmm

    [09:25] azeez: as glen suggested, I;d like to also think abt moving this to AxisEngine [09:25] Deepal: the issue with that is [09:25] Deepal: when the jave return come back to engine then the mesage has gone [09:25] azeez: that wont work for me [09:25] Deepal: I mean jave return come to the Engine.recieve after it call engine.send [09:26] gdaniels: right [09:26] gdaniels: that won't work without changes [09:26] azeez: I will need to send a fault to the client if an error occurs during replication [09:26] Deepal: so what I think is engine is not the right places [09:26] Deepal: to get it work I think we need to do a lot of changes [09:26] Deepal: to API [09:26] Deepal: that I do not like [09:26] gdaniels: yup - AMR it is then [09:27] azeez: good [09:27] Deepal: btw , Glen what is the staus of removing MRs [09:27] Deepal: I sow a JIRA [09:27] Deepal: if we do so , we need to change only two places I guess [09:27] azeez: removing MRs???????? [09:28] Deepal: yes the Async once [09:28] Deepal: its just duplicate code [09:29] gdaniels: I'll do that today [09:29] gdaniels: deprecating the async ones

    [09:29] gdaniels: removing will come later :)

    [09:29] Deepal: ok :) [09:30] Deepal: so Azeez will that work for you [09:30] Deepal: I mean calling that method by AIOMR and AMIM [09:30] azeez: I didnt understand the discussion abt removing MRs? [09:30] azeez: which MRs are gonna be removed? [09:30] amila_suriarachc has joined [09:31] gdaniels: The "async" MRs are all basically exact copies of the regular versions [09:31] Deepal: 2800 [09:31] Deepal: I mean JIRA number [09:31] gdaniels: the threading part (spinning off a new thread to do the business logic) will be moved into AMR [09:31] azeez: aha got it.... [09:32] azeez: a property will control the behavior of the MR [09:32] Deepal: yes [09:32] azeez: it'll work for me

    [09:33] azeez: another issue, since getting an update in the morning, one of my tests "CallUnregisteredServiceTest" is failing [09:33] azeez: its an assertion failure [09:34] azeez: any of u guys encountered this? [09:34] gdaniels: which module is the test in? [09:34] azeez: integration [09:34] gdaniels: integration? [09:34] gdaniels:

    [09:34] srinathp: I am working on https://issues.apache.org/jira/browse/AXIS2-2118 pls let me know if any of

[09:41] azeez: return omEle;

  • [09:41] azeez: } [09:41] azeez: public String echoString(String in) { [09:41] azeez: return in; [09:41] azeez: } [09:41] azeez: public int echoInt(int in) { [09:41] azeez: return in; [09:41] azeez: }

    [09:41] azeez: public String throwAxisFault() throws AxisFault { [09:41] azeez: OMFactory fac = OMAbstractFactory.getOMFactory(); [09:41] azeez: OMElement ele = fac.createOMElement("wso2wsas", fac.createOMNamespace("", ""));

    [09:41] azeez: ele.setText("This is a method which simply throws an AxisFault");

    [09:41] azeez: throw new AxisFault(new QName("wso2wsas_code"),

    [09:41] azeez: "This is a method which simply throws an AxisFault", "wso2_node", [09:41] azeez: "admin", ele); [09:41] azeez: }

    [09:42] azeez: public SimpleBean echoStringArrays(String[] a, String[] b, int c) { [09:42] azeez: /

    [09:42] azeez: SimpleBean bean = new SimpleBean(); [09:42] azeez: bean.setA_r(a); [09:42] azeez: bean.setB_r(b); [09:42] azeez: bean.setC(c); [09:42] azeez: return bean; [09:42] azeez: } [09:42] azeez: }

    [09:42] azeez: the WSDL is about 1KM long :D

    [09:43] azeez: Here is the SimpleBean class

    [09:43] azeez: public class SimpleBean { [09:43] azeez: private String[] a_r; [09:43] azeez: private String[] b_r; [09:43] azeez: private int c; [09:43] azeez: public String[] getA_r() { [09:44] azeez: return a_r; [09:44] azeez: } [09:44] azeez: public void setA_r(String[] a_r) { [09:44] azeez: this.a_r = a_r; [09:44] azeez: } [09:44] azeez: public String[] getB_r() { [09:44] azeez: return b_r; [09:44] azeez: } [09:44] azeez: public void setB_r(String[] b_r) { [09:44] azeez: this.b_r = b_r; [09:44] azeez: } [09:44] azeez: public int getC() { [09:44] azeez: return c; [09:44] azeez: } [09:44] azeez: public void setC(int c) { [09:44] azeez: this.c = c; [09:44] azeez: } [09:44] azeez: } [09:50] Deepal: azeez was it wroking b4 [10:00] azeez: ?wsdl is it? [10:01] azeez: ?wsdl was working well b4 [10:02] Deepal: hmm [10:02] Deepal: there was no any changes in ?wsdl generation in the hackathon [10:03] azeez: can u reproduce the error? [10:04] Deepal: will do [10:07] chamikara has joined [10:10] sandakith has joined [10:47] sandakith has disconnected: Read error: 60 (Operation timed out) [10:52] amila_suriarachc: hi

    [10:52] amila_suriarachc: I am working on https://issues.apache.org/jira/browse/AXIS2-2682 [10:53] amila_suriarachc: do Axis2 support soapHeaderFaults

    [10:53] amila_suriarachc: i.e. <soap:header message="tns:POSubmit" part="ConfigurationHeader" use="literal">

    [10:53] amila_suriarachc: <soap:headerfault message="cfg:ConfigurationFaultMessage"

part="ConfigurationFault" use="literal"/>

  • [10:53] amila_suriarachc: </soap:header> [10:54] {aaron} has joined [10:54] {aaron}: .... hackathon at Indiana University? [10:56] azeez: yep, this is the hackathon at Indiana University [10:57] sandakith has joined [10:59] {aaron}: interesting [10:59] {aaron}: anyway, i have a question...i have axis (1.2.x) behind a reverse proxy and the wsdl it autogenerates for deployed services has the

wrong endpoint address

  • [10:59] {aaron}: is there a way I can change that url? like, set a base url/path somewhere? [11:01] azeez: r u using Axis or Axis2? [11:02] {aaron}: axis [11:03] azeez: hmm... most of us guys here are Axis2 developers [11:06] {aaron}: ok, no problem [11:07] {aaron}: it's interesting because I'm at Cornell and I work with a team that's at IU UITS on Kuali projects [11:07] {aaron}: (not sure if you've heard of that) [11:07] {aaron}: they are working on a services bus amongst other things

    [11:08] {aaron}: gotta go, have fun ;) [11:08] {aaron} has disconnected

    [11:12] azeez: Guys, FaultSerializationTest in the kernel is also failing for me [11:12] azeez: do I need to get an update of Axiom ? [11:12] gdaniels: probably [11:19] saminda has disconnected: Remote closed the connection [11:20] srinathp: yes .. that test only work with most uptodate OM [11:24] ajith has joined [11:33] azeez: Hi Ajith... u @ iu? [11:34] ajith: nope [11:34] ajith: still in cal [11:34] azeez: cool [11:34] ajith: going there tomorrw

    [11:35] gdaniels: your chair is all ready for you here Ajith :) [11:41] davidillsley has disconnected: Read error: 104 (Connection reset by peer) [12:28] Chinthaka has joined [12:33] Deepal: hmm [12:34] Deepal: we are continue on JIRA fixing [12:38] azeez: Guys, I committed my code with the call to replication after invocation of biz logic [12:38] azeez: gotta go now [12:38] Deepal: cool [12:38] Deepal: so is this way , everything working fine for you [12:38] Deepal: or do you need any more changes [12:39] azeez: yes tits fine

    [12:39] azeez: :D [12:39] azeez: sorry I meant "its fine" [13:00] azeez has disconnected: Read error: 113 (No route to host) [13:09] Chinthaka has disconnected: Read error: 104 (Connection reset by peer)

    • Deepal: having the lucnh time discussion on wsdl nampspaces

[13:44] Deepal: handling and relationships

  • [13:52] Deepal: is that ok to store broken wsdl model in wsdl4j [13:52] Deepal: though API allows to do that [13:52] chathura: not clear [13:52] chathura: say tht again

    [13:53] gdaniels: There's a WSDL which just contains <service> in namespace A [13:53] gdaniels: It imports WSDL from namespace B, which contains bindings, portTypes, etc [13:53] gdaniels: So when that gets parsed by WSDL4J we get *two* WSDLDescription objects

    [13:54] gdaniels: (or whatever they're called :)) [13:54] gdaniels: and one has a link to the other [13:54] gdaniels: because each description has only one target namespace [13:54] chathura: got it [13:54] gdaniels: when we're doing codegen, we merge the two into a single WSDL4J structure [13:54] gdaniels: which the API really should not allow us to do [13:54] gdaniels: but it does, apparently [13:55] chathura: now this is why you should have comeup with the componenet model when you proposed wsdl 1.1

    [13:56] chathura: :-) [13:56] gdaniels: lol [13:56] gdaniels: blame Sanjiva

    [13:57] chathura: :-) [14:05] ajith: ok here is my 2 cents [14:05] ajith: WSDL4J is supposed to be only the object model [14:05] ajith: its just another representation of the WSDL file you have [14:06] ajith: if your WSDL is wrong then obviously the object model is wrong ! [14:06] ajith: I beleive this is same in Woden [14:06] ajith: none of them try to validate the structure [14:06] gdaniels: but WSDL has rules [14:07] gdaniels: if my WSDL is targetNamespace="foo", I should not be able to add a portType in namespace "bar" to that description hierarchy [14:07] chathura: think the point is wsdl4j has dropped some of the semantic rules that it need to enforce [14:07] gdaniels: yes exactly [14:07] gdaniels: it shouldn't even be possible to do [14:07] ajith: WSDL4J IMHO should be looked at as a convenient representation of the WSDL XML file you have [14:07] chathura: and the question is should we enforce them [14:07] gdaniels: i.e. I should not be able to add a portType by QName [14:07] gdaniels: I should only be able to add it by NCName [14:07] ajith: yes - exactly [14:08] gdaniels: and the NS should be inferred from the Description I add it to [14:08] ajith: I guess the view taken by both WSDL4J and Woden is that they are mere object represenations of what you have as XML [14:09] ajith: and validations come seperately [14:09] gdaniels: I haven't looked at the APIs recently [14:09] gdaniels: but if you can really break things that easily that seeems not so good [14:09] ajith: so you can write a validator on top of Woden/WSDL4J but they themselves would not validate the WSDL [14:09] chathura: well woden does validation during parsing [14:10] ajith: it does ? [14:10] chathura: yup [14:10] ajith: so if you have a ns missing it complains ? [14:10] chathura: well i think it runs in two modes [14:10] chathura: one with assertion ON mode [14:10] ajith: oh yes - now I remember [14:11] chathura: in that it checks for wsdl 2.0 assertions [14:11] ajith: yes right - but that mode is very slow to operate IIRC [14:12] ajith: in anycase the argument is that what is the scope WSDL object models should cover [14:12] ajith: I mean should they validate or not [14:12] chathura: yup thats where i was heading [14:14] ajith: BTW is this related to the hackathon by any chance ? [14:14] ajith: I mean what we are discussing here [14:31] chathura has disconnected [14:36] chathura has joined

    [15:09] chamikara has disconnected: "ChatZilla [Firefox]" [15:49] chathura: OT: I have a JSON question if someone is willing to give a hand

    [15:49] chathura: <?xml version='1.0' encoding='UTF-8'?>

    [15:49] chathura: <soapenv:Envelope xmlns:soapenv="http://schemas.xmlsoap.org/soap/envelope/">

    [15:49] chathura: <soapenv:Body>

    [15:49] chathura: <ns1:echoString xmlns:ns1="http://ws.apache.org/axis2/xsd">

    [15:49] chathura: <ns1:str>Hello</ns1:str>

    [15:49] chathura: </ns1:echoString>

    [15:49] chathura: </soapenv:Body>

    [15:49] chathura: </soapenv:Envelope> [15:49] chathura: THats the SOAP [15:50] chathura: Here is the JSON

    [15:50] chathura: {"ns1:echoString":{"@xmlns":{"ns1":"http:\/\/ws.apache.org\/axis2\/xsd"},"str":{"$":"Hello"}}} [15:50] chathura: you might notice some namespaces are missing [15:51] chathura: eg str should be ns1:str

[15:56] Deepal: hmm

  • [15:57] Deepal: im sorry I wont be able to answe this [16:02] Deepal: Azeez I think I fixed your issue

    [16:02] Deepal: so now if the service thws AxisFault , we do not add those infor to wsdl [16:06] chathura: wsdl?? [16:07] chathura: i am talking about the runtime [16:07] ajith: BTW I have a q relating directly to Axis2 [16:07] ajith: the build is failing for me [16:07] chathura: i could reuse the adb for json if the above is in the proper format

    [16:07] ajith: Running org.apache.axis2.faults.FaultSerializationTest [16:07] chathura: think i should look at the jsonom serializer [16:11] ajith: chathura : yes [16:11] ajith: BTW this is the trace

    [16:11] ajith: Tests run: 2, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 0.144 sec <<< FAILURE!

    [16:11] ajith: testFaultReason(org.apache.axis2.faults.FaultSerializationTest) Time elapsed: 0.025 sec <<< FAILURE!

    [16:11] ajith: junit.framework.ComparisonFailure: expected:<[myFaultReason]> but was:<[]> [16:11] ajith: I guess this has to do with some new change [16:11] ajith: ? [16:17] Deepal: it is wroking for us

    [16:24] Deepal: http://wiki.apache.org/ws/FrontPage/Axis2/changesfrom1%2e2to1%2e3#preview [16:24] ajith: hmm [16:24] Deepal: please add the changes you have done from 1.2 to 1.3 here [16:24] ajith: you guys using the mvn 2 build ? [16:25] Deepal: so that we can make sure that we include all those into the release note

    [16:25] Deepal: http://wiki.apache.org/ws/FrontPage/Axis2/changesfrom1.2to1.3 [16:25] ajith: sure will do [16:25] Deepal: thx all [16:26] ajith: hmm... was there any changes to axiom ? [16:26] ajith: (not that I've seen any) [16:30] Deepal: not yet [16:30] Deepal: may be we will have [16:37] ajith: I just took a fresh checkout and this issue is there [16:40] chathura has disconnected [16:43] ajith: hmm.. this is strange - It passes inside the ide but fails in the maven build [16:44] Deepal: let me run and see [16:50] ajith: sure - btw what is the mvn version you have

    [16:51] Deepal: hwy how abt SimpleHttpServer.deloyService(classname) [16:51] Deepal: shall we add smt like [16:51] Deepal: one step , you can deploy a service [16:59] Deepal: Ajith : When will you be here [16:59] Deepal: tomorrow mornning ?? [16:59] ajith: tomorrow evening [16:59] Deepal: cool [17:00] ajith: no flights to come in the morning [17:00] ajith: I was planning to leave tonight but there are no flights that fly at night [17:00] ajith: so I have to come early tomorrow morning [17:00] Deepal: cool [17:22] gdaniels: hey cool

FrontPage/Axis2/hackathon_13_06 (last edited 2009-09-20 22:48:48 by localhost)