Re: Large XML file opinions
From: Oleg Tkachenko [MVP] (oleg_at_NO!SPAM!PLEASEtkachenko.com)
Date: 02/09/05
- Next message: chris yoker via DotNetMonster.com: "Re: XSLT "choose" incorrectly assigning the "Test""
- Previous message: Oleg Tkachenko [MVP]: "Re: XSLT "choose" incorrectly assigning the "Test""
- In reply to: Marc Thompson: "Large XML file opinions"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 09 Feb 2005 13:03:47 +0200
Marc Thompson wrote:
> I have several large XML files (500-700 MB) that get updated once a day. I
> have an app that will need to Query (read only operations is all I'll ever
> need) these files. Obviously, querying a file this large will be somewhat
> of a challenge because it can't be loaded into memory all at once. Also, I
> am skeptical of the speed of Xpath/Xquery.
>
> I am wondering if I shouldn't just create a table in SQL server, then create
> a routine that dumps the XML file's values into corresponding SQL server
> columns.
>
> More background info, my app will need to query this XML data fairly
> regularly, about every 5 minutes that the app is being used. So speed and
> efficiency are crucial.
Alternatively you may want take a look at XML data type columns in SQL
Server 2005.
Otherwise you would need to load that XML into XPathDocument (read-only
hence simpler, faster and smaller than XmlDocument) and query it.
Optimize your queries (never use // etc).
If your queries are similar ones, you might try XPath indexing using
IndexingXPathNavigator from Mvp.Xml library (http://mvp-xml.sf.net/common).
-- Oleg Tkachenko [XML MVP, MCP] http://blog.tkachenko.com
- Next message: chris yoker via DotNetMonster.com: "Re: XSLT "choose" incorrectly assigning the "Test""
- Previous message: Oleg Tkachenko [MVP]: "Re: XSLT "choose" incorrectly assigning the "Test""
- In reply to: Marc Thompson: "Large XML file opinions"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
