The second process is when the buffered data is written to the indexes. There are two main sets of parameters that affect this stage—the rate at which the data is written to the indexes (to reduce CPU and disk contention), and the index settings themselves.
Parameters that affect the write process are shown in Table 2-24.
Parameter |
Default |
Description |
|---|---|---|
omniq.indexer.sleepDurationMillis |
20 |
The time, in milliseconds, the indexer thread sleeps during indexing to allow other CPU-intensive applications to run. |
omniq.indexer.sleepFrequency |
20 |
Indicates the number of omniq.indexer.sleepFrequency cycles the indexer thread will sleep. |
omniq.index.term.numSegments |
5 |
The number of segments helps to distribute the indexed data across a number of files, reducing the “seek” times of large files. |
omniq.index.term.minimizationFactor |
20 |
The branching factor of each index segment. This parameter affects the lookup performance of the index segment. |
omniq.index.term.useRootChildrenCache |
true |
If set to true, the index segments cache some of their structure in memoryto improve indexing and querying performance. |
omniq.index.metadata.numSegments |
2 |
The number of segments helps to distribute the indexed data across a number of files, reducing the seek times of large files. |
omniq.index.metadata.minimizationFactor |
10 |
The branching factor of each metadata index segment. This parameter affects the lookup performance of the metadata index segment. |
omniq.index.metadata.useRootChildrenCache |
true |
If set to true, the metadata index segments cache some of their structure in memory to improve indexing and querying performance. |
omniq.lexicon.document.maxKeyLength |
256 |
The maximum document file path length deemed valid for indexing. |
omniq.lexicon.document.minimizationFactor |
20 |
The branching factor of each document lexicon segment. This parameter affects the lookup performance of the document lexicon segment and should not be changed without consulting with technical support. |
omniq.lexicon.document.useRootChildrenCache |
true |
If set to true, the document lexicon segments will cache some of their structure in RAM to improve indexing and querying performance. |
omniq.lexicon.reverseDocument.numSegments |
4 |
The number of segments helps to distribute the indexed data across a number of files, reducing the seek times of large files. |
| Copyright © 2005. Sybase Inc. All rights reserved. |
|
|
