Forum Replies Created
-
Posts
-
You don't need to use SQL Server to find the message parts.
You can use the Group Hub page in the Admin Console to query for in process messages and you can use the HAT tool to view completed messages that have been tracked.
Hi Nar,
I am not able to build the application it gives some error message as "Use of unconstructed message 'msgIncompleteOrder' and msgIncompleteOrder:message has not been initialised in construct statement.Please sen me early reply.
Made everything written in this blog
http://weblogs.asp.net/cschittko/archive/2004/06/07/150582.aspx
But error again appears.
Note: there were no "Windows NT" section in Windows2000 registry. After doing steps written in the bloq above section was created.
HI,
Some amendments to my previous post. I need with Mapping.. for example
i have input file as
FName
LName
Course
FeeI need output file as
FullName (which will combine FName and LName)
Course
FeeThanks
Jen
The Disassemble stage of the pipeline is different from other stages. It works on a first match principal and only executes one component.
What does your custom component do?
That is my thought. I am from the SQL Server world and BizTalk is new to me, so I immediatly think to use DTS/SSIS. But I was trying stay to one technology and take advantage of BizTalk technologies like changing output formats easily and applying business rules. But in the sample runs I have done I see a 9 million row export taking well over a 2-3 hours in time and I have not even applied business rules yet in my orch.
I am running a single high end server, and was starting to wonder if BizTalk just can't handle that kind of volume. If a 100 messages a second is normal processing then I can use that as a base calculation… and show that even in doubling the throughput it still has a performance issue.
10 million records in 3 hours = 925 records per second.
To achieve some sort of parallel processing you will need to split the SQL message in the pipeline, process in multiple orchestrations and append the output records into a file. The receive port and send port will be single threaded, but the orchestrations will multi-thread.
If you require the order of records to be preserved, then you will need to single thread the entire process.
Personally I would look at using SSIS, Biztalk introduces a lot of overhead that you do not need for this process.
You will need to use a custom Xslt.
Here is an example that works with the examples above:
<?xml version="1.0" encoding="UTF-16" ?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" xmlns:msxsl="urn:schemas-microsoft-com:xslt" xmlns:var="http://schemas.microsoft.com/BizTalk/2003/var" xmlns:userCSharp="http://schemas.microsoft.com/BizTalk/2003/userCSharp" exclude-result-prefixes="msxsl var userCSharp" version="1.0">
<xsl:output indent = "yes" omit-xml-declaration="yes" method="xml" version="1.0" />
<xsl:template match="/">
<xsl:apply-templates select="/Root" />
</xsl:template>
<xsl:template match="/Root">
<Root>
<xsl:for-each select="Record">
<Record>
<xsl:if test="@attrib1">
<xsl:attribute name="attrib1">
<xsl:value-of select="@attrib1" />
</xsl:attribute>
</xsl:if>
<xsl:variable name="reset" select="userCSharp:Reset()"/>
<xsl:for-each select="following-sibling::*">
<xsl:if test="userCSharp:MapChild(local-name())">
<NoteToRecord>
<xsl:if test="@attrib1">
<xsl:attribute name="attrib1">
<xsl:value-of select="@attrib1" />
</xsl:attribute>
</xsl:if>
</NoteToRecord>
</xsl:if>
</xsl:for-each>
</Record>
</xsl:for-each>
</Root>
</xsl:template>
<msxsl:script language="C#" implements-prefix="userCSharp">
<![CDATA[
bool done = false;
public void Reset()
{
done = false;
}
public bool MapChild(string nodename)
{
if (nodename == "Record")
{
done = true;
}
return !done;
}
]]>
</msxsl:script>
</xsl:stylesheet>Siva,
1. go and see any errors comming in Event Viewer(BizTalk Administrator- expand Event viewer), click on application.
2. BizTalk 2004->HAT in the wizard Queries->Recent Service Instance
Hope you will find information one of the place.
Nar-
I am still in testing stages on a lab server trying to build an application. Presently I am looking at the possibility (extremely high one at that) of pulling upto 9 Millions records. I will be pulling them from a SQL Server database, applying business rules, and then exporting them as a single file (or set of a dozen broken by a code) to CSV style format.
Is BizTalk built to even come close to pushing 5-10 million records from a SQL Server to a Flat File in say 2-3 hours?
I know there are a lot of unknowns… but in theory is it possible? Are you likely talking 1 server, 10 servers? I will not hold anyone to it… just trying to evaluate and decide if other options than BizTalk should be persued.
Thanks.
