Deduplicate Job - Search Unavailable error

"Search unavailable" is a Salesforce platform error. It can appear while a Deduplicate Job runs within Salesforce, and in most cases the job still completes with reliable results.

Symptoms

• In the Job Overview, the Result column of a Find Duplicates task states First error: Search unavailable, while the Status is Completed.
• Your Salesforce admin may receive Apex error emails from the dupcheck.dc3SearchBatchDelta class, with the message "Search unavailable".
• The error mostly occurs on scheduled jobs that process a large number of records.

Cause

While running a job, Plauti Deduplicate sends search requests to Salesforce to find potential duplicates. "Search unavailable" means Salesforce did not respond to such a request at that moment. This has two common causes:

• Salesforce search was temporarily unavailable. This is a platform-level issue outside of Plauti Deduplicate's control. It usually resolves on its own.
• The job is too large or complex for the Salesforce platform. A job that processes a large number of records, or uses a complex scenario, sends many search requests in a short time. This makes the error more likely to occur, particularly on scheduled jobs that scan the full object on every run.

Impact on job results

If the job's Status is Completed, the job ran to the end and the error was logged along the way. In most cases the results are reliable.

If the error appears only once, no further action is needed. However, if it appears on several runs of the same job or similar jobs, the job's scope is likely the cause rather than Salesforce's search being unavailable.
Apply one or more of the following options to solve this.

Solution

Reduce the load the job puts on Salesforce search. There are several ways to achieve this; start with option 1, since it resolves the error in most cases. 

The options below focus on scheduled jobs. Resolving this error is less urgent for standalone jobs, but you can also apply them there, to improve processing speed.

1. Limit the records each run processes.
   a. Go to the Deduplicate Job tab and switch to the Scheduled Job Overview.
   b. On the job, click Job Options > Edit.
   c. Add or tighten a filter so each run only processes recently created or modified records, for example records created in the last week instead of the last year. These records are still compared against all existing records.
   d. Check that no filter is left incomplete or empty.
2. Simplify the scenario. A scenario with many fuzzy-matched fields requires more search effort. Remove fields that do not add to the match quality.
3. Process the job outside Salesforce. If the job needs to cover a large number of records, process it with Plauti Cloud or Plauti Server instead of Salesforce, as these do not depend on Salesforce search capacity. For standalone jobs, you can also process the job on Plauti Desktop.
   • Costs apply for Plauti Cloud jobs.
   • Scheduled jobs can only run on Plauti Desktop in combination with Plauti Server.