Best practices when data source is a database

Hi ,

I am having a scenario where all the data comes from MS SQL database .
The requirement includes Alarming for which i need to move all the data into tags.

The scale of this data that i need to query from database is what i am worried about .

There are like 20000 individual SQL tags of type Integer, float and string or 2000 SQL tags of type dataset . SQL tags will be randomly queried at interval of 3 to 4 hours sometimes even altogether.

Will this approach affect the performance of the ignition system ? is it even possible ?

Thanks &regards,
Aravindan.S

Something that might make things easier would be to know if there is a timestamp in the database. If there is then you could periodically poll the database to return a list of all the alarms which had events since the last poll and use that to update the tags by script.
The alternative - and possibly easier - would be to poll the database on schedule and update or create tags as the results come in.