January 10, 2020 at 6:31 pm
Plan on creating a database.We will store a Document ID, Document Type. JSON, Path, and Datetime in a master table with indexing in separate tables for each document type. Fields required for indexing will be defined for each type. ie. Member docs will use SBSB_ID, and MEME_SFX.
Audit table to record all api get requests and include Who, Where the request originated, and When
The question is, Deadlock issues are a concern considering how many documents can be added at a time. Support dirty reads on the tables and only lock for inserts. Create a replicated DB for read only
Partition the tables by year would increase performance? Please advise. Thanks in advance!
January 10, 2020 at 9:03 pm
Any thoughts?
January 10, 2020 at 10:55 pm
Also wanted to add to the above, that the master table will have data with pointers to pdf or other document type which are located in the file system. The data would be loaded frequently and we need to read the data with proper indexing so to avoid locking, blocking or deadlocking unless there is any better option other than replication. Instead i was thinking configure replication and read from the replicated data so OLTP will not be impacted.Please advise?
January 11, 2020 at 9:35 pm
Any advise?
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply