Indexing and De-Indexing both run a batch process to include a predefined number of documents at one run. These batch processes are used for performance optimization. If more documents are included in the batch, system performance increases.
The number of documents you can process is limited; the limit depends on the size of the documents and the linked attachments. The document size is calculated based on the pure text and its attachments. Imagine and format elements are not calculated.
Note: You can limit the size of attachments by navigating to Attachments Library, Repositories on the Administration tab and editing the repository to set the File Limit Size (KB).
The recommended batch maximum size is between 2-12 MB (per the EBR_MAX_INDEX_BATCH_SIZE parameter of the NX.env file and the average document size).
@EBR_MAX_INDEX_BATCH_SIZE=128 @NX_EBR_INDEX_QUEUE_TIMEOUT=10 @NX_EBR_REINDEX_QUEUE_TIMEOUT=1 @NX_EBR_INDEX_QUEUE_ONLINE=Yes @NX_EBR_NON_KD_INDEX_QUEUE_ONLINE=Yes
This setting means that one batch processes 128 documents, that batch executions have ten second intervals, and when in reindex, the wait interval between two batches is one second.
@EBR_MAX_INDEX_BATCH_SIZE=25 @NX_EBR_INDEX_QUEUE_TIMEOUT=10 @NX_EBR_REINDEX_QUEUE_TIMEOUT=10 @NX_EBR_INDEX_QUEUE_ONLINE=No @NX_EBR_NON_KD_INDEX_QUEUE_ONLINE=No
This setting means that one batch processes 25 documents, that batch executions have ten second intervals, and when in reindex, the wait interval between two batches is ten seconds.
Copyright © 2012 CA. All rights reserved. | Tell Technical Publications how we can improve this information |