Welcome to the Push Updates Listener for Aspire. A central location for all information on the Aspire Push Updates Listener service.
How It Works
The Push Updates Listener is an Aspire Service that allows external systems to push update requests into an Aspire Content Source. Each request will have one or more documents, each with one of the following actions: add, update and delete. The content source will then start a new incremental crawl and only process the documents that were sent on the request.
All requests will be stacked in a separate queue called pushRequestQueue, and each request will be processed sequentially using "first-in, first-out" policy.
Properties
Property
Required
Description
contentSource
Yes
The name of the content source where documents will be pushed
pushType
No
Values: pushUpdates or cleanAndRestart
Default value when property is not sent is pushUpdates which will send the documents to the pushRequestQueue to be processed by the content source
cleanAndRestart will force release any in progress (status 'P') requests, clean up flags and restart the listener threads.
documents
Yes
Array with all documents that will be pushed
documents.url
Yes
Url or Id of the document
documents.action
Yes
add, update or delete
documents.metadata
No
Any extra metadata required by the content source to process the document
documents.shouldScan
No
Default is false when property is not present. This will force scan the document.
The name of the http servlet that will accept push update requests. The value will appear as a servlet under the root of the server on which Aspire is running. For example, for "updatesListener" your updates listener requests will be handled at http://<server>:<port>/updatesListener.
Wait for Requests
waitRequestsTime
5m
Time to wait for new requests to be pushed after the ones in progress are done. If no new requests are sent and time expires, the crawl will be stopped.
Update Snapshot
pushUpdateSnapshot
false
If enabled, the incoming updates will update the snapshot of the content source. This property only works with Aspire Snapshots. Any connector with custom incremental crawls/snapshots, will ignore this property.
Delete Complete Requests
deleteCompleteRequests
false
If enabled, all requests marked as complete in the queue will be removed on the following crawl start.
Secure servlet
APIKeyEnabled
false
Use an API Key to secure communication with the servlet.
Yes. In this case the node that gets the first request will be in charge of processing all requests sent during the crawl. If that node crashes during the crawl, remaining requests will be processed by any of the other nodes. Document processing is done in parallel by all nodes.
What happens if the content source is disabled?
Any requests sent to the content source while being disabled will get an error response.