September 19, 2012 at 7:10 am
Hi everyone
I was wondering what the best solution/practice would be for the following scenario:
We have a complex data structure with many tables creating one product, there are +- 20 million records. These products get updated on a daily basis and anywhere from 100K - 1 million records are affected each day.
We have a online read only database as well that holds a watered down version of the product record, a basic flat structure of the most crucial fields.
What would be the best way to keep the online version up to date, but have it updated outside of working hours?
I have 2 ideas, but not sure if either are the best or even plausible.
1. Create the update/insert queries that will populate the online version and that will connect form our internal system to the online version and run it as a job each night at a certain time.
2. Setup some sort of replication, Log shipping? I realize that in this option, I will need to create a local version of the online table(s) and populate it with the same queries used in step 1, but then the replication will handle the actual data transfer etc.
Or is there another way of simplifying this task?
Thank you.
September 25, 2012 at 10:42 pm
Have you explored Change Tracking?
There are no special teachers of virtue, because virtue is taught by the whole community.
--Plato
September 25, 2012 at 11:01 pm
Maybe SSIS execute sql task
September 26, 2012 at 12:16 am
ESAT ERKEC (9/25/2012)
Maybe SSIS execute sql task
This gets my vote as what you after is a denormalised version of the data which logshipping and replication is not going to give you (without a lot of development anyway).
With some sort of change tracking, timestamp etc you can update only what has changed as well.
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply