Fixing typo :-)
1) A query with the DATE_ADD function d hit doesn't the index of
TIME_CREATED. ==> 1) A query with the DATE_ADD function doesn't hit the
index of TIME_CREATED and does a full table scan.
On Mon, May 21, 2018 at 8:43 PM Rushmin Fernando wrote:
> Sorry if I
Sorry if I didn't explain it correctly.
It is not the calculation itself. Because the calculation based on fields
of the table the DB is not able to use the indexes for the search.
Please see a simplified test below.
1) A query with the DATE_ADD function d hit doesn't the index of
TIME_CREATED.
On Mon, May 21, 2018 at 8:10 PM, Tharindu Edirisinghe
wrote:
> Currently in the IDN_OAUTH2_ACCESS_TOKEN table, we store the time of token
> generation in TIME_CREATED column in human readable date time format.
> Therefore at the time of token generation, we'll have to get the
Currently in the IDN_OAUTH2_ACCESS_TOKEN table, we store the time of token
generation in TIME_CREATED column in human readable date time format.
Therefore at the time of token generation, we'll have to get the current
timestamp and store it in another column and also add the VALIDITY_PERIOD
to
Hi Devs,
Right now it is the validity period which is getting stored along with the
tokens.
So in order to expire a token, a calculation is needed during expiry tasks.
When there a lot of tokens, this becomes a bottleneck since the database
indexes are not applicable.
But if the expiry