What is the advantage of using the broker over the IS in a pub/sub scenario? Since documents can be published into and out of the IS, how does adding a broker help?

Thanks,
mow

All depends on the architecture that you want to implement.

If you need to route documents between ISs, you can use the broker for this.

But if you have only one IS, and you goal is to make a process asynchronous you can use the IS local publish. In this case, the guaranteed local publish seems to be slower than a guaranteed publish to the broker. But if you use a local volatile publish this will be much faster as the documents reside in the IS memory (this is true from webMethods 6.0.1 SP1).

[url=“wmusers.com”]wmusers.com

Thank you for your feedback on this.

To my understanding, you cannot configure the IS for fail-over & load-balancing (of published documents) when you do a local-publish.

“Since documents can be published into and out of the IS…”

You’re correct that IS can easily move documents around but this isn’t usually referred to as “publishing.” In a pub/sub environment, subscribers register interest in a particular document type with a broker. Publishers send documents to the broker and the broker copies and forwards the document to all registered subscribers. The wM Broker provides this functionality. Without the broker, you’d need to implement your own pub/sub facilities that can communicate outside of the single IS box.

IS has a “local” pub/sub facility. With this facility, documents do not travel out of the IS environment. It’s just a way of moving data around in the IS environment itself. Kinda cool. And as Uday points out, there is no load-balancing or fail-over–which seems obvious since the feature is “local publish/subscribe.”

I am using a Enterprise integrator 5.0.1. I am using a IO adapter to read from file and publish documents . The documents are subscribed by a Oracle adapter. The document types are Volatile with a Time to Live of never discard.
In our production environment I am facing a issue that some of the documents are lost some times. This is verified by means of count mechanism in table.
What are the opportunities that the document is lost. How can I find out which are the documents that are lost.Is there a mechanism to find out that the document has been published but never susbscribed.
There are no Enterprise Server shut down. The memory of the m/c is 4Gb running on Sun Solaris.

Please mail to ssarava8@ford.com or bneelima@ford.com

Document persistence is controlled not only by the doc type definition but also by the queues that the docs travel through. Volatile should only be used when you don’t care if the doc makes it to the destination. This is useful for docs where the data is time sensitive and becomes meaningless after a time, or where subsequent docs will override previous docs (e.g. current temperature). Volatile should normally be used for request/reply ops too.

For volatile docs in volatile queues, these will be lost when the Broker is restarted. There is no mechanism with normal Broker facilities to determine which documents had no subsribers–if a doc has no subscribers, the Broker tosses it away.

It’s been my experience that dropped docs are usually due to the interaction between the adapter/broker client and the resource it interacts with. If you don’t set things up properly there, docs can be lost, though the Broker has successfully delivered it to the adapter.

Hai,

how publish the document into broker through the IS and where can i see that document and where it is stored.

with regards
venkat

Venkat,

Create a publishable document using the Developer tool, and use this service to publish the document to local IS/Broker (pub.publish:publish) and create a Trigger in the IS to subscribe that published document and route it to a flowservice.

The Document stored in the broker queue log to this page ISAdminConsole/WmBrokerAdmin check the link for BrokerServers and browse the links.

Also see the ISAdministrator guide for more information for precise explanation.

HTH,
RMG

Hai,

How to create trigger in IS.

Venkat,

In the Developer tool look in the menu File/New opens a big box with many options like newflowservice, javaservice,documentType, AdapterService etc…and you will also find Trigger option there select that and follow the rest.

Go thru the ISDeveloper guide,it will help for starters.

HTH,
RMG.

Hai,
When i am converting EdI 210 to XML .I am using following services

pub.file:getFile
wm.b2b.edi:converttoValues
pub.xml:documenttoxmlString

I have allready created 210EDI Scheam and its working fine when i am running the above services wm.b2b.edi:converttoValues showing the following error

sun.io.MalformedInputException

Any help is highly Appriciated,
Regards,
VENKAT.

Hai,
When i am converting EdI 210 to XML .I am using following services

pub.file:getFile
wm.b2b.edi:converttoValues
pub.xml:documenttoxmlString

I have allready created 210EDI Scheam and its working fine when i am running the above services wm.b2b.edi:converttoValues showing the following error

sun.io.MalformedInputException

Any help is highly Appriciated,
Regards,
VENKAT.

Venkat,

What version of IS/EDI are you using?

In the ConvertToValues step make sure you are providing a valid edidata string and EDIFFSchema.

“MalformedInputException” indicates that some duplicate input is passing or pipeline messing up.

Also look in the Integration Server & Trading Networks/ EDI for webMethods section,here many queries related to WM EDI stuff,

So before posting your problem search in that section group,it saves time.

Goodluck,

hai,

Iam publishing document into broker,then subscribing that document using one trigger.After trigger receive the document i give one service that is define in the trigger service but this service is not run.Tell me what is the problem.

with regards,
venkat

Venkat,

Are you seeing any error in the logs?Always give us more details so that response will be close to resolve.

Also make sure your subscribing service input is the publishable documentName (set fully qalified document name)example:folder.subfolder:documentName(RecordReference) this will extract the data from the publishable document.

So for debugging purpose in the service first step use savePipelineToFile and later restore it check the pipeline.you should see the result as expected.

HTH,
RMG.

Venkat,
See to it that the subscribing service (the one you are calling from the trigger) has in the input the fully qualified name of the document you are publishing. See that you have used the fully qualified name.

Regards,
Pradeep.

Hi, I have published a doc to the IS itself. But when i subscribing the doc using a Trigger and flow service, the contents of the doc is NULL, that is after subscrbing there is no contents in the doc. Can any one give me some idea on this?

Hi siwa,

U did in correct process. After publishing the document go to service which you mentioned in trigger and check out pipeline data by using save pipeline and restore pipeline data. whenever u restore the pipeline data, in result panel u can findout document which came from broker. just copy that document name and save in input tab of the service. Now you can subscribe the data. all the best and take care.

Hi BSBabu!, No . This is my intention. I have a doc. I am validating, and publshing to the IS. I have trigger which subscribe the published doc to a Service.That service needs to convert that doc into XML string for further processing. So am just converting that doc into XML string in the subscribing service.So u know, the documentTOXML needs “Document” as the input. So while tracing that line, exception is raised as

Exception: java.lang.Exception: missing required parameter ‘document’ while executing trigger. Rejecting Document for TriggerStore:XML:Trigger.

But i have mapped the pipeline IN 's document (which is the Doc ref to the publishable document).

Please tell me what is the problem here.

BSBabu! can u reply me soon?

I am SiWa and just now registered with wmusers. So SiWa = SivaRaman. So pls i need reply soon . Thanks BSBabu.

Hi All!.
This is my intention. I have a doc. I am validating, and publshing to the IS. I have trigger which subscribe the published doc to a Service.That service needs to convert that doc into XML string for further processing. So am just converting that doc into XML string in the subscribing service.So u know, the documentTOXML needs “Document” as the input. So while tracing that line, exception is raised as

Exception: java.lang.Exception: missing required parameter ‘document’ while executing trigger. Rejecting Document for TriggerStore:XML:Trigger.

But i have mapped the pipeline IN 's document (which is the Doc ref to the publishable document).

Please tell me what is the problem here.

Siva,

The input to your triggered service needs to be the fully qualified name of your publishable document. If not, the trigger will place data into your service’s pipeline, but not in the document you are expecting. This causes your documentToXMLString service to fail because the document you are mapping to its “document” input parameter is null.

The quickest way to see what the trigger is actually passing to your service is to disable all steps in your flow service and add a single pub.flow:tracePipeline statement as the first line. This will output the contents of the pipeline to your server’s log file or to the console (depending on how you started IS). You can also use pub.flow:savePipelineToFile as BSBabu suggested.

Your service input tabe must contain a document whose name matches exactly the (fully qualified) name of the document that the trigger is passing.

HTH,

Mark

Hi, Mark. thanks. I have done like that . I got the subscribed contents in the file written using pub:savePipelineToFile . But i have given the correct qualified path in the Trigger as u said. But why i am not able to convert that document into XML string . Note, i have got the XML content in the File Using pub:savePipelineToFile . Any reason for that. Thanks to BSBabu.

Siva,

In the subscribe service the service input should be (fully qualified publishable document name syntax is folder.subfolder:documentName(RecordReferenc).Then only the xml content which is in the pipeline will be extracted.

HTH,
RMG