On Fri, Apr 17, 2020 at 8:28 AM Ted Toth <txt...@gmail.com> wrote: > > > On Thu, Apr 16, 2020 at 8:09 PM Rob Sargent <robjsarg...@gmail.com> wrote: > >> On 4/16/20 6:15 PM, Adrian Klaver wrote: >> > On 4/16/20 4:59 PM, Ted Toth wrote: >> >> >> >> >> >> On Thu, Apr 16, 2020 at 6:29 PM Ted Toth <txt...@gmail.com >> >> <mailto:txt...@gmail.com>> wrote: >> >> >> >> I've noticed that the first exec of an INSERT prepared statement >> >> takes ~5 time longer (I'm using libpq in C and wrapping the calls >> to >> >> time them) then subsequent exec's is this the expected behavior and >> >> if so is there any thing I can do to mitigate this affect? >> >> >> >> Ted >> >> >> >> >> >> For example (in my environment) I'm seeing the prepare take ~10ms, >> >> the first exec take ~30 ms and subsequent exec's take ~4 ms. >> >> >> > >> > I don't have an answer. I believe though that to help those that might >> > it would be helpful to show the actual code. >> > >> > >> You expect the subsequent calls to benefit from the cached query parse >> and planning. What does you query cost without begin wrapped in a >> prepared statement (preferably from a cold start). >> >> >> >> I thought that's what the PQprepare call was supposed to do i.e. > parsing/planning. > > It's a bit difficult to get an unprepared query cost since there are a lot > of columns :( > #define INSERT_SQL "INSERT INTO t (<column names>) VALUES > ($1,$2,$3,$4,$5,$6,$7,$8,$9,$10,$11,$12,$13,$14,$15,$16,$17,$18,$19,$20,$21,$22,$23,$24,$25,$26,$27,$28,$29,$30,$31,$32,$33,$34,$35,$36,$37,$38,$39,$40,$41,$42,$43,$44,$45,$46,$47,$48,$49,$50,$51,$52,$53,$54,$55,$56,$57,$58,$59,$60,$61,$62,$63,$64,$65,$66,$67,$68,$69,$70,$71,$72,$73,$74,$75,$76,$77,$78,$79,$80,$81,$82,$83,$84,$85,$86,$87);" > > Ah prepare does the parsing and execute does the planning.
Another related question is that my implementation uses strings for all values is there a performance benefit to using actual values and specifying their datatype?