I don’t think it was tongue in cheek.
This doesn’t look very ETL, it looks TEL as you are doing the transform in
the source environment. The whole point of an ETL is to reduce the workload
on your source environment, this isn’t doing that.
A good solution for the overall problem could look more like:
* Incremental changes from source database (only read what’s changed and
what will make a difference to what you have on your target system).
* Staging area that calculates what needs changing based on the source
changes
* Load these changes to the target system.
The staging area could be on your target system.
There are many products you can purchase to make this go smoothly but it
will depend on your budget and amount of effort you’re willing to make.
This is where an outside consultant can really help you decide (yes, you’ll
end up paying a lot but if they have the expertise and can properly
determine what your needs are then you are going to be in a much better
place).
Thanks,
Andy
On Mon, 31 Jan 2022 at 20:27, Mladen Gogala <gogala.mladen@xxxxxxxxx>
wrote:
On 1/31/22 05:39, Jonathan Lewis wrote:
1) Upgrade to 19c
2) Hire someone to sign an NDA and get them to do a proper database design
after a full briefing on requirements.
You are NOT asking for a "design suggestion", you're asking for yet
another hack/stop-gap to keep your system limping along until it gets so
unwieldy that it will be virtually impossible to do anything to fix it.
Regards
Jonathan Lewis
Two points:
1. It is usually I who writes tongue in cheek posts like this one.
2. What you are describing is the usual point of application redesign.
When the design gets too messed up to do anything that makes sense, usually
the hardware is thrown at the problem until the redesign is done.
Regards
--
Mladen Gogala
Database Consultant
Tel: (347) 321-1217https://dbwhisperer.wordpress.com
-- //www.freelists.org/webpage/oracle-l
