I have a table in mind which basicaly would look like this:
table1 -------- id bigserial autoincrement primary key, token1 varchar(32) unique not null, token2 varchar(32) unique not null, token3 varchar(32) unique not null [........some other data ......]
The tokens are UUIDs, just saved as varchar strings and without dashes.
My client web app is build is such a way that I can’t always use “id” to retrieve a record. Namely, I need to be able to retrieve a record by one of the “tokens” too.
Also, my client web app may need to update the “token(s)” also, different ones at different times, if a certain condition ocurrs. Not too often.
I estimate that the table would grow, at most, at 100.000 records per day.
1) Will using these 3 tokens, of 32 characters long, as oppossed to using only “id”,
2) Will updating them these
worsen the performance significantly?