OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: storing xml files into database



Haresh:

 

Our Extensible Information Server (XIS) native XML data server product was the first native XML data server product to market back in March of 1999.  We have over 350 customers including SwissRe Insurance, AXA Insurance, Amazon.com, NTT DoCoMo, The Associates, and Laing Construction.  We have invested thousands of man years invested in building a highly scalable and fast native XML data server and have been involved in the XML standards groups since early 1997.  Our internal testing and our own customers’ testing of using RDBMS systems to store XML have shown not only severe functional limitations but severe limitations in terms of scalability and performance as well.  Our internal tests using repositories of 1,000 to 100,000 XML documents about 10KB each in size show that we are over 4 times faster than the RDBMS systems in bulk document loading and over 6 times faster on both indexed and non-indexed cross-document queries.  Results we have seen from our own customers show similar results.  We have attempted tests using repositories of 1 million to 10 million documents and the RDBMS systems consistently fail at those tests.  They cannot complete the benchmark.  We have tested our product with over 10 million documents (resulting in database sizes in the hundreds of gigabytes) and exhibit linear scalability from 1 to 10 million (and, based on customer feedback, even beyond 10 million).  We can make all of these results available to you.

 

Our product is very easy to use as it presents a file system-like interface for managing document storage.  Our system is a fully ACID-compliant transactional database with support for incremental updates (using our own UpdateGram syntax) and a patented distributed caching and locking architecture that insures consistently fast access to data as applications scale.  Our newest release coming out the end of this month also supports distributed X/A transactions.  We have full support for XML 1.0 with Namespaces, XPath 1.0, and XSLT 1.0 (XSLT transformations run right in the server and are highly scalable… our internal XSLT engine is up to 10x faster than standalone XSLT processors) and validation support against both DTDs and XML Schemas.  We support structural, value, and word indexing at the node level with support for single document as well as cross-document XPath queries.

 

You can download a 30-day evaluation edition our product from our website at www.exceloncorp.com.  It is a fully functional version (not crippled like most of our competitors ;->), so you can actually run our benchmarking tests against the evaluation version.  The Extensible Information Server (XIS) is part of the eXcelon XML Platform.  The download link is on our front page under “Download”.  If you need an extension to your evaluation, let me know at any time by contacting me directly.

 

If you have any more questions regarding the eXcelon products or would like more technical information, including benchmarking data or access to our benchmarking test code from our Benchmarking Center, do not hesitate to contact me directly.

 

Cheers,

Chris

 

---------------------------------------

Chris Parkerson

Product Manager

eXcelon Corporation

Burlington, MA

(781) 674-5393

http://www.exceloncorp.com

---------------------------------------

 

-----Original Message-----
From: Haresh Gujarathi [mailto:gharesh@vsnl.com]
Sent: Sunday, September 09, 2001 8:20 AM
To: Xml-Dev
Subject: storing xml files into database

 

I am developing an internet based application where the the data of the users of that application is stored in xml format.

I expect about 1000 to 20000 total users of the system (and say about 100 users simultaneously accessing the system)

 

Naturally there is inconvenience in storing the xmls on the file system. I want to explore storing these xmls in the database. I need some inputs as to how the xmls can be stored in the rdbms database. Please provide me any insight regarding this.

 

I am looking for 2 solutions ; a zero cost solution and reasonable cost solutions (may be involving the xml aware databases like Ixiasoft xml server)

 

1. In considering a database as storage mechanism, what advantages I would get

2. In such scenario what other people do?

3. If I want to store the xmls in say MySQL, how do I do it and does mysql have any special features?

4. What about MS-SQL server;

5. Should xmls be stored as blobs in the rdbms databases?

 

6. Optionally, I am also looking for 'search' ability where the database should be able to index the elements of xml file and perform search.

 

regards

Haresh Gujarathi