Stream: implementers
Topic: Resource Identification, within an SQL RDBMS
Dimitar (Nov 03 2017 at 16:02):
Hello, I was wondering, what would be the best approach to deal with resource identification, within an SQL RDBMS, where resources are stored as a JSON formatted blob. Would it make sense to identify it, based on its full External URL or is there a more efficient and smarter way to identify the target resource, locally? Thank you in advance!
Grahame Grieve (Nov 03 2017 at 20:34):
the logical id is intended for this use
Dimitar (Nov 03 2017 at 21:02):
So it is supposed to be resolved, through the NamingSystem?
Dimitar (Nov 04 2017 at 10:40):
In my case, I am dealing with extraction from multiple different servers and I need to re-identify the resources.
Grahame Grieve (Nov 04 2017 at 12:29):
then I'm not sure exactly what you are asking
Dimitar (Nov 04 2017 at 14:53):
@Grahame Grieve My use case is the following, I am retrieving patient records, from multiple different servers and storing them. Now my concern is, how should I uniquely identify the resource to be stored, so that it can be updated, if it already exists in the local database. I am currently doing it by storing the fullUrl of a resource as a primary key and doing a lookup on all of the ExternalUrls, to match the retrieved record, but I feel that this approach is highly inefficient and wanted to see if there is a better way of doing this. I am trying to achieve something like the following
Client dealing with multiple servers paragraph.
Grahame Grieve (Nov 05 2017 at 19:38):
well, logical ids are 64 characters so that you can generate compose ids. That can save you doing lookups - which sounds like the inefficient bit
Dimitar (Nov 05 2017 at 22:47):
Excuse my ignorance, but what do you mean by generating compose ids?
Grahame Grieve (Nov 05 2017 at 22:48):
well, an id is a set of characters and digits, and - or . which can be up to 64 characters
Grahame Grieve (Nov 05 2017 at 22:49):
you could prepend [x] - where [x] is a code for the source system, and just use the source system's logical identifier
Dimitar (Nov 05 2017 at 22:53):
Aha, therefore as an example, I have resources coming from Server A and Server B. I would just store the resource, retrieved from server A as ServerA/Observation/[logical id]? And do the lookup on the code for source system, resource type and logical id?
Last updated: Apr 12 2022 at 19:14 UTC
