Optimization of a large database query - mysql

Question

I would like to ask if there is any way to optimize the following query:
SELECT SQL_BIG_RESULT DISTINCT u.spy_tech,
u.computer_tech,
u.military_tech,
u.defence_tech,
u.shield_tech,
u.energy_tech,
u.hyperspace_tech,
u.combustion_tech,
u.impulse_motor_tech,
u.hyperspace_motor_tech,
u.laser_tech,
u.ionic_tech,
u.buster_tech,
u.intergalactic_tech,
u.expedition_tech,
u.metal_proc_tech,
u.crystal_proc_tech,
u.deuterium_proc_tech,
u.technologia_bankowa,
u.technologia_kolonizacyjna,
u.technologia_optymalizacji,
u.technologia_otchlani,
u.graviton_tech,
SUM(p.small_ship_cargo) as small_ship_cargo,
SUM(p.big_ship_cargo) as big_ship_cargo,
SUM(p.light_hunter) as light_hunter,
SUM(p.heavy_hunter) as heavy_hunter,
SUM(p.crusher) as crusher,
SUM(p.battle_ship) as battle_ship,
SUM(p.colonizer) as colonizer,
SUM(p.recycler) as recycler,
SUM(p.spy_sonde) as spy_sonde,
SUM(p.bomber_ship) as bomber_ship,
SUM(p.solar_satelit) as solar_satelit,
SUM(p.destructor) as destructor,
SUM(p.dearth_star) as dearth_star,
SUM(p.battleship) as battleship,
SUM(p.lune_noir) as lune_noir,
SUM(p.ev_transporter) as ev_transporter,
SUM(p.star_crasher) as star_crasher,
SUM(p.giga_recykler) as giga_recykler,
SUM(p.dm_ship) as dm_ship,
SUM(p.apocalypse) as apocalypse,
SUM(p.annihilator) as annihilator,
SUM(p.devastator) as devastator,
SUM(p.fleet_save) as fleet_save,
u.r_badawczy,
u.r_budowlany,
u.r_militarny,
u.r_obrony,
u.r_paliwowy,
u.r_gospodarczy,
u.r_planetarny,
u.r_magazynowy,
u.r_energetyczny,
u.r_ekonomiczny,
u.r_odkrywczy,
u.r_max,
u.r_speed,
u.r_moon,
u.r_dm,
u.r_terra,
SUM(p.misil_launcher) as misil_launcher,
SUM(p.small_laser) as small_laser,
SUM(p.big_laser) as big_laser,
SUM(p.gauss_canyon) as gauss_canyon,
SUM(p.ionic_canyon) as ionic_canyon,
SUM(p.buster_canyon) as buster_canyon,
SUM(p.small_protection_shield) as small_protection_shield,
SUM(p.big_protection_shield) as big_protection_shield,
SUM(p.planet_protector) as planet_protector,
SUM(p.graviton_canyon) as graviton_canyon,
SUM(p.orbital_station) as orbital_station,
SUM(p.dzialo_pulsacyjne) as dzialo_pulsacyjne,
SUM(p.dzialo_fotonowe) as dzialo_fotonowe,
SUM(p.interceptor_misil) as interceptor_misil,
SUM(p.interplanetary_misil) as interplanetary_misil,
u.id,
u.ally_id,
u.onlinetime,
s.tech_rank AS old_tech_rank,
s.build_rank AS old_build_rank,
s.defs_rank AS old_defs_rank,
s.fleet_rank AS old_fleet_rank,
s.total_rank AS old_total_rank,
s.total_points_old AS total_points_old,
s.fleet_points_old AS fleet_points_old,
s.defs_points_old AS defs_points_old,
s.build_points_old AS build_points_old,
s.tech_points_old AS tech_points_old
FROM test_users as u
LEFT JOIN test_statpoints as s ON s.stat_type = 1 AND s.id_owner = u.id
LEFT JOIN test_planets as p ON u.id = p.id_owner
WHERE u.user_lastip != "127.0.0.1"
AND u.bana = "0"
AND u.authlevel = "0"
AND u.universe = "1"
GROUP BY s.id_owner, u.id
On my VPS, the query takes 8 seconds and the result is about 10,000 results. Although I improved the machine twice, to 8x2,4CPU, 16gb RAM, the query time has not changed.
So I thought to optimize the above-mentioned query or work on VPS configuration and find a bottleneck that can generate such time.
Table STATPOINTS:
Table USERS:
I do not know if I have given you an insight into the tables, when it comes to something else I would ask for guidance.
Explain:

Answer 1

Your query can run faster if you add the following indexes:
create index ix1 on user (bana, authlevel, universe);
create index ix2 on test_statpoints (stat_type, id_owner);
create index ix3 on test_planets (id_owner);
They include the access columns according to the join predicates.