On 7/15/20 9:38 AM, Tom Lane wrote:
> Adam Sanchez <a(dot)sanchez75(at)gmail(dot)com> writes:
>> I need to run 2 million queries against a three columns table t
>> (s,p,o) which size is 10 billions rows. The data type of each column
>> is string. The server has 512G RAM, 32 cores and 14T SSD (RAID 0)
>
>> Only two types of queries:
>
>> select s p o from t where s = param
>> select s p o from t where o = param
>
> TBH, this is a pretty silly way to use a SQL database. Put
> the probe values into a temporary table (in batches, perhaps)
> and do a join. The per-row cost of that sort of approach
> will be multiple orders of magnitude smaller than a query
> per row.
>
> regards, tom lane
>
>
I had assumed the OP was planning a performance test for say 2M separate
user/requests (perhaps non-unique) rather than asking for the specific
2M records.