On Thu, Feb 15, 2018 at 5:58 AM, Kyotaro HORIGUCHI <horiguchi.kyot...@lab.ntt.co.jp> wrote: > It is amost a just-to-delete work but I see two issues raised so > far.
I think the two issues you are pointing out are the same issue -- namely, if initdb probes for a max_connections setting with an invalid DSM implementation configured, it will fail the test for every value of max_connections and thus select the lowest possible value. The solution to this is presumably just to make sure we choose the DSM implementation before we do the max_connections probing so that we can pass through the correct value there, which I think is what your patch does. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company