Sunday, April 3, 2016

Endeca Basics : How to troubleshoot partial update is working as expected or not?

Endeca Partial indexing process uses only Forge process to generate XML file and Dgidx process gets skipped.   The MDEX Engine processes p... thumbnail 1 summary
Endeca Partial indexing process uses only Forge process to generate XML file and Dgidx process gets skipped.  

The MDEX Engine processes partial updates concurrently with processing incoming queries. This function is also known as the continuous query. The partial update does not require restarting of Dgraph servers.


Find out steps below how partial gets applied to MDEX(Multi Dimensional Index)

1. Once XML gets generated by Forge process, MDEX engine process updates on Record-by-record basis.

2. Generated XML from step 1 copied to <<Endeca/App>>/data/dgraphs/DgraphA1/dgraph_input/updates folder.

3. Update fails or succeeded at the record level.

4. File gets deleted under <<Endeca/App>>/data/dgraphs/DgraphA1/dgraph_input/updates folder once partial update gets applied to MDEX.



If there any error occurs during partial update :-

1. Partial update process continues to apply to MDEX.

2. A log entry added to Dgraph logs

3. MDEX engine writes a  file containing the failed updated record and dimension value under <<Endeca/App>>/data/dgraphs/DgraphA1/dgraph_input/updates/failed_update folder

Click here to get more topics related to Endeca basic concepts.

No comments

Post a Comment

Text Widget