community-content

Forum Replies Created

Viewing 15 posts - 1,651 through 1,665 (of 9,648 total)
  • Author
    Posts
  • Thanks a lot this gives me god idea ..however forum plz ad any more considerations if possible.

    Firstly the WCF SQL adapter. While the hashkey table idea may provide some level of comfort it will kill any scalability. You are basically single threading the the database access. Only the first host instance will ever work, as all other queries will die with a concurrency violation until the first query has completed ( i.e. when the message is written to the message box). The UPDLOCK, READPAST locking hints are a proven technique for creating a re-entrant queue in SQL – here is one of many articles written on the subject http://www.mssqltips.com/tip.asp?tip=1257

    Secondly, why do you output the SQL result to a file and then read it back in again. Why not bypass this step and have the subscriber of the file adapter, subscribe to the WCF SQL adapter.

     

    Your receive location must be resident in IIS/WAS on a BizTalk server. The receive location code is responsible for receiving the message from the wire, running the pipeline, any maps and writing to the MessageBox database. If this code executes in the DMZ you will need to open firewall ports to expose the BizTalk SQL databases to the DMZ. This is less secure than opening ports 80, 443 to allow HTTP access to a web service inside your network. 

    The standard practice for offering public web services from BizTalk is to use a reverse proxy such as ISA server or create your own web service in the DMZ that calls the BizTalk web service internally. My preferred option is the reverse proxy approach. The ISA server or other vendors product  is specifically built as a security device. You do not have to write code and be concerned about potential security issues in this code. The reverse proxy is less maintenance and probably more secure.

    Latest ISA server is now called Forefront Threat Management Gateway 2010

    in reply to: Can’t retrieve all data from soap exception #24188

    I solved this by using Biztalk 2006 R2 and WCF,  in the send port I disabled propagate fault message. Now the full detail field is readable in the exception.

     

    Regards,

     

    Bart

    in reply to: Object Reference not set to an instance of an object #24187

    Initilization of the .Net class must be done however ..also ensure the scope of the variable if you initilize it some where else…and use it might not work so make sure its a Orcehstartion level variable.

    Hi Scott,

     

      Use the BizTalk Administration Console to troubleshoot why this error is happening, you should see a suspended instance of your Orchestration, and look for the message type schema that is causing this to fail.

      The error message “Send Port: “WcfSendPort_SqlAdapterBinding_TypedProcedures_dbo_CrawCustom” URI: “mssql://.//Claims_Risk_Management_Staging?InboundId=CrawConnection” should give you a clue: are you missing the schema that binds to your sql stored procedure (generated by the WCF adapter)?

     

       Daniel.

     

    in reply to: Object Reference not set to an instance of an object #24184

    Hi,

     

      If you did not declare your C# class and methods as static, then you will have to instantiate your class object in the BizTalk Expression shape, before invoking your method. Do something like this:

     

    varMyClass = new Class();

    varString = varMyClass.classmethod(str);

     

       Daniel.

     

    in reply to: WCF-SQL Adapter #24183

    You must create a variable of xmlDocument type. For example:

    xmlDoc.LoadXml(@”ns0:Response xmlns:ns0=’http://WebServiceSimulado.schRetorno’>

    <MSG_RETORNO>OK</MSG_RETORNO>

    </ns0:Response>”);

     

    msgIn_Resp = xmlDoc;

    The xmlDoc variable is xmlDocument type.

     

    in reply to: Object Reference not set to an instance of an object #24181

    Hi Thanks for the response, here is my code, the method returns a string.

    ResponseString = varValidateXml.ValidateXml(ResponseString);

    varValidateXml is the variabe that I have created ti instantiate the class.

    Thanks

    in reply to: Object Reference not set to an instance of an object #24180

    Do you return a variable string in your .NET assembly?

    Type your code here, so we can analyze better.

    OK, that worked. I can now test the map. I have now published it BizTalk, created my send and receive ports, but i’m back to the same error when I try and run a file through:

    There was a failure executing the response(receive) pipeline: “CrawfordTestProject.ReceivePipeline1, CrawfordTestProject, Version=1.0.0.0, Culture=neutral, PublicKeyToken=2a1607f4a050e392” Source: “Spreadsheet disassembler (Spread for BizTalk)” Send Port: “WcfSendPort_SqlAdapterBinding_TypedProcedures_dbo_CrawCustom” URI: “mssql://.//Claims_Risk_Management_Staging?InboundId=CrawConnection” Reason: Parsing the incoming stream failed because Spread for BizTalk was unable to find a schema that matched the format of the incoming stream.

     I cannot seem to find why I am getting this error. This can’t be THAT difficult…

    Ok, the image did not show up in my previous post, so click on my avatar and look for it in My Files link, the name of the image is Map.png

     

    I hope this helps you troubleshoot your Map testing and validation…

     

    Hi Scott,

     

       Why don’t you use the xml data you generated from step 1 as your input file to the map. In VS, in Solution Explorer, right-click on your map file and choose Properties, and select that XML file you generated in the TestMap Input Instance field (as shown below):

     

       Daniel.

     

    HI Daniel,

    I do appreciate your responses.

    This is long before I even get to BizTalk. I have spent they walking through this step by step to see where it might be, so let me see if I can explain my process.

    1) Using the  Spreadsheet schema wizard, create a schema based on the .xls file. Once created, I validate the schema then validate the instance. Both validate successfully. The resulting xml looks like this:

    <id_id>2</id_id>

      <id_client_id_client>15113</id_client_id_client>
      <id_claim_no_id_claim_no>0105-073757-001</id_claim_no_id_claim_no>
      <id_claimid_id_claimid>SIS 151130105073757001</id_claimid_id_claimid>
      <id_brfile_id_brfile>0105-073757</id_brfile_id_brfile>
      <clmt_name_clmt_name>ANDRADE REIGOBER</clmt_name_clmt_name>
      <clmt_age_clmt_age>23</clmt_age_clmt_age>
      <clmt_wage_clmt_wage>520</clmt_wage_clmt_wage>
      <clmt_len_clmt_len>12</clmt_len_clmt_len>
      <cov_major_cov_major>WC</cov_major_cov_major>

     2) Create a pipeline, using the Spreadhseet disassemblere and setting the Document Schema to the schema I created in step 1.

     3) Add an Adapter Service (Generated Item) that creates a schema for my insert stored procedure.

    4) Add a Map. The source schema is my file schema and the Destination schema is the stored procedure schema. I then map a few string columns and save the map.

    5) I then validate the map. This is successful.

    6) I then TEST the map. This fails. I look at the file used for the test map, and it looks like this:

      <ns0:id_id>10</ns0:id_id>
      <ns0:id_client_id_client>10</ns0:id_client_id_client>
      <ns0:id_claim_no_id_claim_no>id_claim_no_id_claim_no_0</ns0:id_claim_no_id_claim_no>
      <ns0:id_claimid_id_claimid>id_claimid_id_claimid_0</ns0:id_claimid_id_claimid>
      <ns0:id_brfile_id_brfile>id_brfile_id_brfile_0</ns0:id_brfile_id_brfile>
      <ns0:clmt_name_clmt_name>clmt_name_clmt_name_0</ns0:clmt_name_clmt_name>
      <ns0:clmt_age_clmt_age>10</ns0:clmt_age_clmt_age>
      <ns0:clmt_wage_clmt_wage>10</ns0:clmt_wage_clmt_wage>
      <ns0:clmt_len_clmt_len>10</ns0:clmt_len_clmt_len>
      <ns0:cov_major_cov_major>cov_major_cov_major_0</ns0:cov_major_cov_major>

    Why are the values for the elements the NAME of the element and not the actual values? This is why I am getting the error i received earlier:

    Output validation error: The ‘http://schemas.microsoft.com/Sql/2008/05/TypedProcedures/dbo:line_of_coverage_type_code_value&#8217; element is invalid – The value ‘cov_major_cov_major_0’ is invalid according to its datatype ‘String’ – The actual length is greater than the MaxLength value.

    cov_major is only 10 characters long. But because it is bringing in the element name as the value, it is failing.

    I am lost at this point. BizTalk isn’t even in the picture yet and i need to find out why the map test is failing, or more specifically, why it is using the element names for the string values and the number 10 for numerical values.

    Hi Scott,

     

       Sorry for not responding sooner but it’s a hectic week for me at work…

       A good sample of input XML file for your map to test and validate would be to get the XML message coming from your Initial Receive port (after your original Excel data has been parsed into XML format by the Spread pipeline). You can generate such a message by placing a Send port in your Orchestration after the initial Receive port and declare its type of your Spread schema, build and deploy your map project, then in the Administration console, create a new port of with a Transport Type of File and a Send Pipeline type of XMLTransmit.

     

       Daniel.

     

Viewing 15 posts - 1,651 through 1,665 (of 9,648 total)