[BUG] OpenSearch source has expiring PIT

Describe the bug

I'm seeing issues in the opensearch source that indicate that the PIT is expiring while reading from an index.

2025-12-11T07:59:49.545 [pool-16-thread-1] ERROR org.opensearch.dataprepper.plugins.source.opensearch.worker.PitWorker - Unknown exception while processing index 'my_index':
org.opensearch.client.opensearch._types.OpenSearchException: Request failed: [illegal_argument_exception] Point In Time id: [UJZuozGIhKWjjjWIHGayhg] doesn't exist
 at org.opensearch.client.transport.aws.AwsSdk2Transport.extractAndWrapCause(AwsSdk2Transport.java:720)
 at org.opensearch.client.transport.aws.AwsSdk2Transport.performRequest(AwsSdk2Transport.java:222)
 at org.opensearch.client.opensearch.OpenSearchClient.search(OpenSearchClient.java:1386)
 at org.opensearch.dataprepper.plugins.source.opensearch.worker.client.OpenSearchAccessor.searchWithSearchAfter(OpenSearchAccessor.java:270)
 at org.opensearch.dataprepper.plugins.source.opensearch.worker.client.OpenSearchAccessor.searchWithPit(OpenSearchAccessor.java:132)
 at org.opensearch.dataprepper.plugins.source.opensearch.worker.PitWorker.processIndex(PitWorker.java:189)
 at org.opensearch.dataprepper.plugins.source.opensearch.worker.PitWorker.lambda$run$0(PitWorker.java:115)
 at io.micrometer.core.instrument.composite.CompositeTimer.record(CompositeTimer.java:141)
 at org.opensearch.dataprepper.plugins.source.opensearch.worker.PitWorker.run(PitWorker.java:115)
 at org.opensearch.dataprepper.plugins.source.opensearch.OpenSearchService.lambda$start$0(OpenSearchService.java:108)
 at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:539)
 at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264)
 at java.base/java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:304)
 at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136)
 at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635)
 at java.base/java.lang.Thread.run(Thread.java:840)

To Reproduce

I'm not sure how to reproduce this.

Expected behavior

I believe there should be two things:

1. There seems to be a bug with how we extend the PIT timeout.
2. Pipeline authors should be able to configure the PIT timeouts to prevent this.

Screenshots

N/A

Environment (please complete the following information):

Data Prepper 2.13.0

Additional context

N/A

Start with OpenSearchAccessor.searchWithPit and searchWithSearchAfter, then inspect PitWorker.processIndex at the stack-trace locations. Trace how the PIT timeout is extended and how failures are handled, and determine where pipeline configuration could supply the timeout. Done means the PIT remains usable during reads and authors can configure its timeout, with tests covering the behavior.