Export (0) Print
Expand All

Registry keys reference (Search Server 2008)

Search Server 2008

Applies To: Microsoft Search Server 2008

 

Topic Last Modified: 2009-08-26

NoteNote:
Unless otherwise noted, the information in this article applies to both Microsoft Search Server 2008 and Microsoft Search Server 2008 Express.

Microsoft Search Server 2008 uses registry values to store many configuration values. You can set some of these values in the Central Administration site and you should do so when you can. For example, you should set the ConnectTimeout, DataTimeout and IgnoreCertCNError values by making changes in the Central Administration\Application Management\Search Service\Farm-Level Search Settings page. However, many values can only be set in the registry by using Registry Editor. If you change any of the registry values listed in this article, you must restart the Office SharePoint Server Search service for the changes to take effect. However, if you make registry value changes while the Office SharePoint Server Search service is not running, the changes will be overwritten when the service starts.

WarningWarning:
Do not change any Microsoft Search Server 2008 registry values or keys that are not listed in this article unless advised to do this by Microsoft Support.
CautionCaution:
Incorrectly editing the registry may severely damage your system. Before making changes to the registry, you should back up any valued data on the computer. For more information about the registry see the following article: Windows registry information for advanced users (http://go.microsoft.com/fwlink/?LinkID=142984).

You can find the following values in this registry location:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Office Server\12.0\Search\Global\Gathering Manager

 

Name Default Value User Interface Location Description

ByPassList

Blank

Search Server\Search Administration\Farm-Level Search Settings

If the network uses a proxy server for access to the Internet, you can configure the name of that server in the Central Administration site. If you do this, Search Server uses the proxy server during crawls. You can also configure a list of addresses for which the proxy server is not used. This list is stored in the ByPassList registry value. See also the LocalByPassProxy value.

ConnectTimeout

60

Search Server\Search Administration\Farm-Level Search Settings

The time in seconds that the index server waits for a response when it connects to a content source during a crawl. If a content source responds slowly, its content may not be crawled. If you increase this value above 60 seconds, slow content sources are more likely to be crawled but the crawl may take longer. To accelerate crawls, decrease this value. However,, be aware that slow content sources may not be crawled and the index may become out of date.

DataTimeout

60

Search Server\Search Administration\Farm-Level Search Settings

The time in seconds that the index server waits for data after it connects to a content source during a crawl. Content sources can respond to connection requests quickly, and still be slow to return data. You can reduce this value to less than 60 seconds to prevent such content sources from slowing the crawl process. If you do this, be aware that, as with the previous value, that the index may become out of date with slow content sources.

DebugFilters

0

Not available

This value is used by IFilter and Protocol Handler developers when they debug their code. When the value is set to 1 or 2, the indexer will not kill the process if it is slow to complete so that developers have time to step through their code. This value should always be set to 0 in production environments.

DebugWordBreakers

0

Not available

This value is used by Word Breaker developers when they debug their code. When the value is set to 1, developers can easily see the words output by their Word Breaker. This value should always be set to 0 in production environments.

DedicatedFilterProcessMemoryQuota

104857600

Not available

This value places a limit on the memory use of search daemons as described below under FilterProcessMemoryQuota. This value is for single-threaded search daemons.

DeleteOnErrorInterval

100

Not available

After an item is added to the index it is checked at each later crawl. If an error occurs when this check happens, the item remains in the index. However, if the number of consecutive errors exceeds this value the item is removed from the index.

FilterProcessMemoryQuota

104857600

Not available

If a search daemon, such as an IFilter or Protocol Handler, uses an amount of memory that exceeds this value, the crawler kills the process. This limit prevents memory leaks in search daemons. The default value is designed for index servers that have 4 GB of physical memory. If you have more memory in the index server, you can increase this value proportionately to cache more data during crawls. However, because only 64 bit systems can use more than 4 GB of memory, do not increase this value unless you have 64 bit hardware and a 64 bit build of Microsoft Search Server 2008. This value is for multi-threaded search daemons.

FlushLinksCount

200

Not available

When the indexer crawls a document, it caches links found in the document in memory or on the disk. If the number of links exceeds this value, the links are flushed from the cache. See also LinkInMemoryCacheSize.

FlushLinksRollback

0

Not available

This value is not used.

FolderHighPriority

50

Not available

This value limits the number of high-priority folders that the indexer can process at the same time. If the value is set too high, the cache in the daemons will constantly be running out of space. If the value is too low, the crawl will be throttled waiting for more items to process. If you increase this value, you should also increase the FilterProcessMemoryQuota and DedicatedFilterProcessMemoryQuota values, because daemons use more memory, so otherwise the crawler kills them frequently and causes crawl errors.

FullIncrementalInterval

10080

Not available

This value only applies when the indexer crawls content that is stored in Microsoft Office SharePoint Portal Server 2003 or Windows SharePoint Services 2.0.This value defines a limit, in minutes, above which an incremental crawl becomes a full incremental crawl. Such a crawl processes some changes, such as deletes, that are not processed by ordinary incremental crawls.

IgnoreCertCNError

0

Search Server\Search Administration\Farm-Level Search Settings

When this value is set to 0, sites secured with Secure Sockets Layer (SSL) encryption will be crawled only if there are no problems with the certificate. The following example scenarios can cause a certificate error:

  • The certificate has expired.

  • The certification authority that issued the certificate is not trusted by the index server.

  • The server name of the content source does not match that in the certificate.

In each of these examples, the connection between the indexer and the content source will still be encrypted so that, if you trust the source, you can choose to allow connections anyway.

LinksInMemoryCacheSize

100

Not available

When the indexer crawls a document it caches in memory links found in the document. If the number of links in the document exceeds this number, the links are cached on the disk. If the index server has a large amount of memory, you should consider increasing this value for faster indexing.

LocalByPassProxy

0

Search Server\Search Administration\Farm-Level Search Settings

If this value is set to 1, the indexer connects directly to content sources in the local network. If this value is set to 0, and a proxy server is configured, the proxy server is used even for intranet addresses.

LocalStoreAccountSecret

Blank

Not available

This value stores an encrypted password for authentication against the local content store. This value should not be changed in the registry.

MaxAnchorsHttp

80

Not available

When the indexer crawls a Web page, it stores the hyperlinks in the page and later crawls those pages. You can control this behavior when you create a content source. For example, you can specify a maximum depth of links to follow. Web pages with very large numbers of links can slow down the crawl process. Therefore, this registry value limits the number of links crawled from a page. If you have many pages that have lots of links, and have lots of time to complete crawls, increase this value.

MaxAnchorsNonHttp

1000000000

Not available

This value is the equivalent of MaxAnchorsHttp for any document that is not a Web page.

MaxGrowFactor

4

Not available

The indexer uses this value to limit the size of a file that it indexes. Protocol Handlers report the size of each file, but some files, such as dynamically generated Web pages, may be significantly larger than estimated. If the file size is larger than the reported file size multiplied by the MaxGrowFactor value, the indexer raises an error and the file will not be processed.

PerformanceLevel

5

Central Administration\Operations\Services on Server\Office SharePoint Server Search Service Settings

This value governs the priority given to the search processes. For index and query servers, this value should be 5, which corresponds to “Maximum” in the user interface. For servers that are not dedicated to search functions, use lower values.

PlugInPingFrequency

12

Not available

This values indicates the number of index service heartbeats between pings to the plug-ins. Plug-ins include IFilters and Protocol Handlers. If you have an unreliable plug-in you should consider reducing this value.

ProxyName

Blank

Search Server\Search Administration\Farm-Level Search Settings

The proxy server name that you enter in the user interface is stored in this registry value. If you enter a port number, the ProxyName value is http://servername:portnumber.

RobotsTxtRefreshFrequency

86400

Not available

Many Web sites use a robots.txt file to list directories and files that should not be indexed. Search Server 2008 respects these preferences, but caches the robots.txt file for the time specified in this value in seconds. The default value of 86400 seconds is 24 hours. If you index sites that regularly change the robots.txt file, you should consider reducing this value.

RobotThreadsNumber

0

Not available

This value limits the number of robot threads that the indexer creates during a crawl. The value 0 is the default and means that the indexer automatically chooses the number of threads. If you find that crawling is using too many resources, such as CPU time or network time, you should consider specifying a value. The maximum value is 64.

UseProxy

0

Search Server\Search Administration\Farm-Level Search Settings

If the network uses a proxy server for access to the Internet, you can configure the name of that server in the Central Administration site. When you specify a proxy server in the user interface, this value is set automatically to 1. If you use Registry Editor to specify a proxy server, make sure that you manually set this value to 1.

You can find the following value in this registry location:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Shared Tools\Web Server Extensions\12.0\Search\Global\Gathering Manager

 

Name Default Value User Interface Location Description

MaxDocId

4294967295

Not available

You can use this value to place an upper limit on the total number of documents that the indexer will crawl. The value -1 means that there is no limit. If the index file is becoming too large, this is one way to limit its size. Note, however, that some documents in the content may never be indexed if you use this value.

Was this page helpful?
(1500 characters remaining)
Thank you for your feedback
Show:
© 2014 Microsoft