INVALID_DOMAIN_LIST (Invalid Domain List) error in crawl, at startup or on the runtime (Doc ID 1040315.1)

Last updated on AUGUST 28, 2017

Applies to:

Oracle Knowledge - Version 8.1.2.1 to 8.4.7.1 [Release 8.1 to 8.4]
Information in this document applies to any platform.
The INVALID_DOMAIN_LIST error can happen at different points in the Inquira indexer, workclient and runtime processes. In general it means that the working copy of the dictionary, found in the userdata\runtime directory, has files that are not readable or are corrupted or missing. This can happen for a number of reasons.
- A bad save to the dictionary from the workbench.
- A bad synch - some files did not get transfered correctly to the runtime.
- A bad propagation, either bad dictionary files from the previous environment or a bad checkout to the working copy of the dictionary in userdata.

These problems have not been seen in 8.5 and above at this point.

Symptoms

Document indexing fails with error: FORWARD_INDEX_PROCESS_DOCUMENT fails with com.inquira.userdata.UserDataException: INVALID_DOMAIN_LIST null

An examination of the application log showed that nearly every page presented for indexing was failing with the following error:

[117 Local Workclient-default-taskrun(639)] Event(Code=FORWARD_INDEX_PROCESS_DOCUMENT, id=#9B0I639P-1) started at 06/10/08 11:58: Indexing document [id: 50, Fetch URL http://localhost/manual/howto/ssi.html] in Collection ApacheManual.
[118 Local Workclient-default-taskrun (640)] Failed to index with exception: com.inquira.userdata.UserDataException: [Local Workclient-default-taskrun @ Mon Oct 06 11:58:48 BST 2008 - Indexing] (INVALID_DOMAIN_LIST null)
    at com.inquira.userdata.domain.impl.DomainDataServiceImpl.getDomainListByName(DomainDataServiceImpl.java:147)
    at com.inquira.userdata.domain.impl.DomainDataServiceClientImpl.findAll(DomainDataServiceClientImpl.java:147)
    etc.

Such errors only caused the indexing of the particular document to fail, not complete aborting of the job. So job processing continued on until global maintenance at which point the job would fail completely. Examination of the log showed the reason for the failure as:

[26 Local Workclient-default-taskrun (3120)] Status: Global maintenance aborted, since there are no active documents!
[27 Local Workclient-default-taskrun(3122)] Event(Code=NO_ACTIVE_DOCUMENTS, id=#11B0I3122P-1) with Failure occurred at 06/10/08 12:07: There are no active documents and no content to be searched.Consequence : Check that you have acquired and properly processed content before continuing.Corrective Action : Check that you have acquired and properly processed content before continuing.
[28 Local Workclient-default-taskrun(3120)] Event(Code=RECEIVE_EVENT, id=#11B0I3120P-1) with Failure ended at 06/10/08 12:07 after 513 ms: Received event 76b6d788-9394-11dd-9a9f-904c16481716#11B1I3119P3118 from scheduler to Local Workclient.
[29 Scheduler(85)] SCHEDULER_RUN_JOB resumed at 01/01/70 00:59: Job 'ApacheManual_FullProcess'.
[30 Scheduler(3118)] SCHEDULER_RUN_TASK resumed at 01/01/70 00:59: Task [ApacheManual_FullProcess/Global Maintenance] assigned to workclient Local Workclient.
[31 Scheduler(3123)] Event(Code=SCHEDULER_TASK_FAILURE, id=#11B0I3123P-1) with Failure occurred at 06/10/08 12:07: Task [ApacheManual_FullProcess/Global Maintenance] completed with errors.
[32 Scheduler(3118)] Event(Code=SCHEDULER_RUN_TASK, id=#11B0I3118P-1) with Failure ended at 06/10/08 12:07 after 579 ms: Task [ApacheManual_FullProcess/Global Maintenance] assigned to workclient Local Workclient.
[33 Scheduler (85)] UNKNOWN MESSAGE ID "SCHEDULER_STALLED_DUE_TO_FAILURE" called with: "com.inquira.scheduler.job.TaskSummary@1d8c1b8"
[34 Scheduler(85)] Event(Code=SCHEDULER_RUN_JOB, id=#6B0I85P-1) with Failure ended at 06/10/08 12:07 after 694,986 ms: Job 'ApacheManual_FullProcess'.
[35 Scheduler (12)] UNKNOWN MESSAGE ID "SCHEDULER_STALLED_JOB" called with: "ApacheManual_FullProcess"
[36 Scheduler(3124)] Event(Code=SCHEDULER_JOB_STALLED, id=#11B0I3124P-1) with Failure occurred at 06/10/08 12:07: Job stalled [ApacheManual_FullProcess].  Due to errors/dependencies, job can no longer make progress.

Hence global maintenance aborted as there was nothing to process because all of the crawled pages had failed indexation.

Changes

You will need to trace back to see what is working.

If the runtime failed with the invalid domain error, see if the last crawl was successful in the system manager, and the indexer ui.jsp is working as expected.

To check the crawl it can be neccessary to look at the indexer binary log for the invalid_domain_list error.  For instructions see (Doc ID 1363089.1)

To check the indexer ui.jsp use this url (http://localhost:8222/inquirawb/ui.jsp), remembering this is a public interface, no security is in place so only public documents will return. 

Indexed collection counts in the system manager can also be used to know if the crawl worked successfully.

If the indexer seems good then simply reset the runtime.  see step 7 in this knowledge article: (Doc ID 2009225.1)

If the indexer is not working or the crawl was not successful check if a propagation was done in the last couple days from the preceding environment.  Then check the indexer in the proceeding environment for the same information about the last crawl, collections and if the ui.jsp is working on the indexer and the runtime.

You have to work your way back to what is working and move that forward again.

You could just need to repropagate or you could have to go back to the dictionary and resave a change and propagate all the way through again.

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms