I am debating using the SQL Server 2000 database or the file system to store large objects (image datatype) and would like to find some discussion about the reading/writing techniques when the files are 300-500 MB. Most of the time the files will be a few MB to tens of MB; however, some could be hundreds of MB.
Once the file is read from the database or filestream, it will be sent via a TCP .NET Remoting service to another TCP .NET Remoting Service to reconstitute the file on the other end. There is really no need to reconstitute the file before it is sent across the internet if it is retrieved from the database. But, with files that large, I'm not sure what technique to use to read the data from the database, much like you would use a FileStream object to read an array of bytes from the file system. I'm wondering if the SqlDataReader would be viable. One problem is that the "Read" method appears to be row based, in which case the entire row could be 500MB. If anyone has any experience with this or articles you could recommend, I would be grateful. Thanks, Todd Hickerson (Also posted on the OLEDB list)
