优化SQL语句（离散连接和左连接）CakePhp

Question

So, I have cakephp SQL statement like this

$groupId = $this->request->data['group_id'];
        $placesT = TableRegistry::get('Places');
        $AllPrices = $placesT->find('all')
            ->select(['Places.postal', 'Places.city', 'Places.district', 'Places.state', 'Places.country', 'Prices.delivery_price', 'Prices.retrieval_price', 'Prices.maintenance_price',
                'Prices.edm_price', 'Prices.weighting_price', 'Prices.price_type_fixed_price', 'Prices.price_type_weight_price_per_ton', 'Prices.fee_per_day'])
            ->leftJoin(
                ['Prices' => 'prices'],
                [
                    'Prices.group_id' => $groupId,
                    'Prices.product_id' => $id,
                    'Prices.valid_to >' => (time() * 1000),
                    'Prices.postal = Places.postal',
                ]
                , ['Prices.group_id' => 'integer', 'Prices.product_id' => 'integer', 'Prices.valid_to' => 'integer', 'Prices.postal' => 'integer'])
            ->distinct()
            ->orderAsc('Places.postal')
            ->toArray();

In SQL looks like:

SELECT DISTINCT places.postal AS `Places__postal`,
                places.city AS `Places__city`,
                places.district AS `Places__district`,
                places.state AS `Places__state`,
                places.country AS `Places__country`,
                prices.delivery_price AS `Prices__delivery_price`,
                prices.retrieval_price AS `Prices__retrieval_price`,
                prices.maintenance_price AS `Prices__maintenance_price`,
                prices.edm_price AS `Prices__edm_price`,
                prices.weighting_price AS `Prices__weighting_price`,
                prices.price_type_fixed_price AS `Prices__price_type_fixed_price`,
                prices.price_type_weight_price_per_ton AS `Prices__price_type_weight_price_per_ton`,
                prices.fee_per_day AS `Prices__fee_per_day`
FROM places
LEFT JOIN prices ON (prices.group_id = '3'
                            AND prices.product_id = '1'
                            AND prices.valid_to > 1488186767000
                            AND prices.postal = places.postal)
ORDER BY places.postal ASC;

This statement works, but when I have bigger data amount (for testing) it doesnt work (takes too long to execute).

Prices table have arount 1M rows, and Places Table arount 3k data.

Is there any way how can I make this statement faster?

Answer 1

Should not use multiple conditions on JOIN. Better to have them on Where condition. Also make sure you have indexed columns like postal and product_id to see better performance.

SELECT DISTINCT places.postal AS `Places__postal`,
                places.city AS `Places__city`,
                places.district AS `Places__district`,
                places.state AS `Places__state`,
                places.country AS `Places__country`,
                prices.delivery_price AS `Prices__delivery_price`,
                prices.retrieval_price AS `Prices__retrieval_price`,
                prices.maintenance_price AS `Prices__maintenance_price`,
                prices.edm_price AS `Prices__edm_price`,
                prices.weighting_price AS `Prices__weighting_price`,
                prices.price_type_fixed_price AS `Prices__price_type_fixed_price`,
                prices.price_type_weight_price_per_ton AS `Prices__price_type_weight_price_per_ton`,
                prices.fee_per_day AS `Prices__fee_per_day`
FROM places
LEFT JOIN prices ON prices.postal = places.postal
WHERE prices.group_id = '3'
   AND prices.product_id = '1'
   AND prices.valid_to > 1488186767000
ORDER BY places.postal ASC;