|
database
newsgroups
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Inserting single record files into SQL 2K5 based on file's existenAX and until we have our AX Returns Subsystem up and running, we want to create a process that if files containing one record each appear in a folder, we want SQL 2005 to insert them into a given table. I have already setup a similar process with an SQL Server Agent job utilizing xp_fileexists as one step, a bulk insert statement as second step, and an xp_cmdshell deleting the source file as a third step. My concerns with this process are as follows: 1.) You can only setup an Agent Job to run at minimum every 1 minute every day. I'd like to have a process that is triggered by a file's existence rather than looking every minute to see if a file exists. My reasoning lies within concerns of performance and processor time being eaten by a process that polls endlessly every 60 seconds. 2.) As I alluded to above, I am concerned with performance issues and whether or not this is the best method (using Agent Jobs) because we could potentially have 30-50 separate Agent Jobs running polling directories for different types of records to insert into different tables. 3.) As far as I can tell, I can't setup the Agent Job to pull in filenames with wildcards. I'd like to be able to pull in multiple files based upon a similar filename like RECORD_FILE_yyyymmdd_hhnnss.csv, where I'd specify the input file as 'RECORD_FILE_*'. Overall my first question is if using SQL Server Agent Jobs given then scenario above is the best method of achieving what I am wanting and if so, how I can supply the xp_fileexists, BULK INSERT, and xp_cmdshell a provision for being able to act upon files of different names where the difference equals the date/time stamp in the filename, and I can make sure that when I delete a file it deletes the one it just inserted into the table rather than one it hasn't inserted yet. Thanks in Advance, David |
|||||||||||||||||||||||