Hi Balduin,
On 07/02/2024 11:05, Balduin Landolt wrote:
Hi everyone,
we're storing data in Fuseki as a persistence for our application backend,
the data is structured according to the application logic. Whenever
something changes in our application logic, we have to do a database
migration, so that the data conforms to the updated model.
Our current solution to that is very home-spun, not exactly stable and
comes with a lot of downtime, so we try to avoid it whenever possible.
If I understand correctly, this is a schema change requiring the data to
change.
The transformation of the data to the updated data model could be done
offline, that would reduce downtime. If the data is being continuously
updated, that's harder because the offline copy will get out of step
with the live data.
How often does the data change (not due to application logic changes)?
I'm now looking into how this could be improved in the future. My double
question is:
1) is there any tooling I missed, to help with this process? (In SQL world
for example, there are out of the box solutions for that.)
2) and if not, more broadly, does anyone have any hints on how I could best
go about this?
Do you have a concrete example of such a change? maybe chnage-in-place
is possible but that depends on w=howupdates happen, how the dada feeds
change with the application logic change.
Andy
Thanks in advance!
Balduin
