XML or database...?
-
I'm having 500 hard coded records in a database. Now I thinking about to make a XML for that..So Which one is faster...? A XML file or a database table ..? If a XML is faster then what is the best way to retrieve the data from it...? A dataset or make a XML parser for that...? :^)
Regards Pankaj Joshi If you want to shape your dreams into reality, please wake-up...
If you store all your data in one single xml-file you will have to query wich data to show by using xpath (I guess you wont show all data at once). Another way is to create some xsl-files. But use a database if you can. It's much easier to work with data stored in a database and I think it's faster too. If this data is static and performance is a big issue you should consider generating static html-files (if it's a webapplication)
--------------- www.serverside.no
-
I'm having 500 hard coded records in a database. Now I thinking about to make a XML for that..So Which one is faster...? A XML file or a database table ..? If a XML is faster then what is the best way to retrieve the data from it...? A dataset or make a XML parser for that...? :^)
Regards Pankaj Joshi If you want to shape your dreams into reality, please wake-up...
Database, ie SQL CE. Faster or not depends on how you use it. Do you load the XML into memory once and use the data from memory? or do you have to query the XML every time? If you load them into memory, you will be hogging a lot of memory. You won't know if you wake up one day, the requirement has been changed to 6000 records, must be able to add or delete, and changes needs to be deployed to 500 users.
-
If you store all your data in one single xml-file you will have to query wich data to show by using xpath (I guess you wont show all data at once). Another way is to create some xsl-files. But use a database if you can. It's much easier to work with data stored in a database and I think it's faster too. If this data is static and performance is a big issue you should consider generating static html-files (if it's a webapplication)
--------------- www.serverside.no
Serverside wrote:
If you store all your data in one single xml-file you will have to query wich data to show by using xpath...
Not really. If you use the class XmlDataDocument, you can manipulate XML data through a relational Dataset, so in a similar manner to that used in regular databases. If there are only 500 records I would recommend either using XmlDataDocument or using an Access database file. The need for a database server like SQL Server or MySql is not justified in such a scenario.
-
Database, ie SQL CE. Faster or not depends on how you use it. Do you load the XML into memory once and use the data from memory? or do you have to query the XML every time? If you load them into memory, you will be hogging a lot of memory. You won't know if you wake up one day, the requirement has been changed to 6000 records, must be able to add or delete, and changes needs to be deployed to 500 users.
darkelv wrote:
If you load them into memory, you will be hogging a lot of memory.
How do you know? He hasn't said what the 500 records are! It might be 500 IDs with a string value consisting of a few words (lets say 20 chars average). That is 24 bytes (or 44 for Unicode). So that is a whopping 22K. Wow! That's a real memory hungry application.
darkelv wrote:
You won't know if you wake up one day, the requirement has been changed to 6000 records, must be able to add or delete, and changes needs to be deployed to 500 users.
Where did he say he had 500 users? And why couldn't the XML file be located in a central location on a network? You are making a lot of assumptions that were not in the original query. If you are going to do that then you should say that these are your assumptions and therefore they may possibly not apply to the situation.
Upcoming FREE developer events: * Developer Day Scotland My website | blog
-
Serverside wrote:
If you store all your data in one single xml-file you will have to query wich data to show by using xpath...
Not really. If you use the class XmlDataDocument, you can manipulate XML data through a relational Dataset, so in a similar manner to that used in regular databases. If there are only 500 records I would recommend either using XmlDataDocument or using an Access database file. The need for a database server like SQL Server or MySql is not justified in such a scenario.
blackjack2150 wrote:
If there are only 500 records I would recommend either using XmlDataDocument or using an Access database file. The need for a database server like SQL Server or MySql is not justified in such a scenario.
What if SQL Server is being used for other data in the application? In that case it becomes a valid candidate. Adding Access to the mix in that scenario would be a very poor idea.
Upcoming FREE developer events: * Developer Day Scotland My website | blog
-
I'm having 500 hard coded records in a database. Now I thinking about to make a XML for that..So Which one is faster...? A XML file or a database table ..? If a XML is faster then what is the best way to retrieve the data from it...? A dataset or make a XML parser for that...? :^)
Regards Pankaj Joshi If you want to shape your dreams into reality, please wake-up...
Pankaj - Joshi wrote:
I'm having 500 hard coded records in a database. Now I thinking about to make a XML for that..So Which one is faster...? A XML file or a database table ..?
It depends on what you are using these records for. How do you intend to query them? What do they contain?
Pankaj - Joshi wrote:
If a XML is faster then what is the best way to retrieve the data from it...?
For example, if the data is a simple key/value lookup then loading it into a Hashtable may be the best thing to do. However, it really depends on what this data is.
Upcoming FREE developer events: * Developer Day Scotland My website | blog
-
Pankaj - Joshi wrote:
I'm having 500 hard coded records in a database. Now I thinking about to make a XML for that..So Which one is faster...? A XML file or a database table ..?
It depends on what you are using these records for. How do you intend to query them? What do they contain?
Pankaj - Joshi wrote:
If a XML is faster then what is the best way to retrieve the data from it...?
For example, if the data is a simple key/value lookup then loading it into a Hashtable may be the best thing to do. However, it really depends on what this data is.
Upcoming FREE developer events: * Developer Day Scotland My website | blog
XML is very faster than Database :)
Like a good handsome...thanking You
-
XML is very faster than Database :)
Like a good handsome...thanking You
anandakumarm wrote:
XML is very faster than Database
That is a very wide sweeping statement that is incorrect in most scenarios.
Upcoming FREE developer events: * Developer Day Scotland My website | blog
-
darkelv wrote:
If you load them into memory, you will be hogging a lot of memory.
How do you know? He hasn't said what the 500 records are! It might be 500 IDs with a string value consisting of a few words (lets say 20 chars average). That is 24 bytes (or 44 for Unicode). So that is a whopping 22K. Wow! That's a real memory hungry application.
darkelv wrote:
You won't know if you wake up one day, the requirement has been changed to 6000 records, must be able to add or delete, and changes needs to be deployed to 500 users.
Where did he say he had 500 users? And why couldn't the XML file be located in a central location on a network? You are making a lot of assumptions that were not in the original query. If you are going to do that then you should say that these are your assumptions and therefore they may possibly not apply to the situation.
Upcoming FREE developer events: * Developer Day Scotland My website | blog
-
Feeling cranky, eh. That's what the "if" for. Since he didn't supply more information, I am giving him something to ponder before he decide what he want to do.
darkelv wrote:
That's what the "if" for.
No, it wasn't. You said: "If you load them into memory, you will be hogging a lot of memory." That indicates that you have already made the assumption that the data set is large, not IF the data set is large. I pointed out a fairly reasonable scenario where the data set wouldn't really take that much memory.
Upcoming FREE developer events: * Developer Day Scotland My website | blog