为 Prestashop 1.6 优化 MySQL

Question

我在具有 4GB RAM、2CPU、60GB SSD 特性的 VPS 上运行基于 Prestashop 1.6 的电子商店。目前，我的商店中有大约 20000 种产品，并且由于 mysql 查询运行时间长，我在加载网站时遇到了问题。当我运行 htop 来分析进程时，我看到 mysql 消耗了两个 CPU 的 100%。这是 mysqltuner 的输出：

-------- Performance Metrics -------------------------------------------------
[--] Up for: 1h 29m 9s (241K q [45.109 qps], 319 conn, TX: 318M, RX: 126M)
[--] Reads / Writes: 78% / 22%
[--] Total buffers: 192.0M global + 2.7M per thread (151 max threads)
[OK] Maximum possible memory usage: 597.8M (15% of installed RAM)
[OK] Slow queries: 0% (8/241K)
[OK] Highest usage of available connections: 2% (4/151)
[OK] Key buffer size / total MyISAM indexes: 16.0M/44.9M
[OK] Key buffer hit rate: 99.6% (36M cached / 133K reads)
[OK] Query cache efficiency: 49.6% (101K cached / 205K selects)
[!!] Query cache prunes per day: 1386761
[OK] Sorts requiring temporary tables: 0% (1 temp sorts / 7K sorts)
[!!] Joins performed without indexes: 78
[OK] Temporary tables created on disk: 8% (846 on disk / 9K total)
[OK] Thread cache hit rate: 98% (4 created / 319 connections)
[!!] Table cache hit rate: 10% (340 open / 3K opened)
[OK] Open file limit used: 62% (643/1K)
[OK] Table locks acquired immediately: 100% (239K immediate / 239K locks)

-------- Recommendations -----------------------------------------------------
General recommendations:
    Add skip-innodb to MySQL configuration to disable InnoDB
    MySQL started within last 24 hours - recommendations may be inaccurate
    Enable the slow query log to troubleshoot bad queries
    Adjust your join queries to always utilize indexes
    Increase table_cache gradually to avoid file descriptor limits
Variables to adjust:
    query_cache_size (> 16M)
    join_buffer_size (> 128.0K, or always use indexes with joins)
    table_cache (> 400)

请提出任何优化方法。

编辑：

慢查询输出日志为here。

Answer 1

这是您日志中的第二个查询，正在查看其他查询。为了帮助解决这个问题，我稍微重组以摆脱子查询并变成 INNER JOIN。此外，为了帮助不了解您的表格的其他人，请始终将您的字段引用限定为 table.column 或 alias.column，例如了解您的 level_depth、nleft、nright 来自哪里。它可能有助于索引。

对于索引，每个表我都会有以下内容。

ps_category, index ON( active, id_category, id_lang, nleft, nright, level_depth )
ps_category_lang, index ON( id_category, id_shop, id_lang )
ps_category_shop, index ON( id_category, id_shop )  
ps_category_group, index ON( id_category, id_group )
ps_lang, index ON( id_lang, active )

稍微重写的查询

SELECT 
      c.id_parent, 
      c.id_category, 
      cl.name, 
      cl.description, 
      cl.link_rewrite
   FROM 
      ps_category c
         INNER JOIN ps_category_lang cl 
            ON c.id_category = cl.id_category 
            AND cl.id_shop = 1
            AND cl.id_lang = 2 
         INNER JOIN ps_category_shop cs
            ON c.id_category = cs.id_category 
            AND cs.id_shop = 1
         INNER JOIN ps_category_group psg
            ON c.id_category = psg.id_category
            AND psg.id_group = 1
   WHERE 
      (  c.active = 1 
      OR c.id_category = 2)
      AND c.id_category != 1
      AND level_depth <= 7
      AND nleft >= 350 AND nright <= 351
   ORDER BY 
      level_depth ASC, 
      cs.position ASC;

第三次查询。请注意，在这种情况下，某些特定的关键元素被移到顶部，因此我也将它们放在键索引中，然后将其余部分放在 where 子句中（主要是可读性调整，但有助于查看“特定”元素指数优势）

SELECT 
      c.id_category, 
      cl.name, 
      cl.link_rewrite
   FROM 
      ps_category c
         LEFT JOIN ps_category_lang cl 
            ON c.id_category = cl.id_category 
            AND cl.id_shop = 1
         INNER JOIN ps_category_shop category_shop
            ON c.id_category = category_shop.id_category 
            AND category_shop.id_shop = 1
   WHERE 
          c.active = 1
      AND cl.id_lang = 2
      AND c.nleft between 2 and 350
      AND c.nright between 351 and 625
      AND c.level_depth > 1
   ORDER BY 
      c.level_depth ASC;

您需要确认您真正想要的下一个查询...您对语言表有一个左连接，但随后在 where 子句中添加“AND l.active = 1”，这实际上会变成它进入 INNER JOIN。如果你真的想要一个 LEFT-JOIN，将 l.active 移动到我在这里调整的连接部分

SELECT 
      l.id_lang, 
      c.link_rewrite
   FROM 
      ps_category_lang AS c
         LEFT JOIN ps_lang AS l 
            ON c.id_lang = l.id_lang
            AND l.active = 1
   WHERE 
      c.id_category = 324

希望这些索引和查询的示例说明/可读性可能有助于改进您的日志。如果其他人仍有问题，请根据需要发布。

对慢速日志报告中的第一个查询进行修订

在第一个查询中，您似乎正在从特定产品商店中寻找商品。但是，您从产品类别开始查询，然后左连接到产品，然后再到产品商店。由于这些产品最终将通过 where 与特定类别相关联，因此无论如何它都会创建一个 INNER JOIN。我会稍微重组如下

我会在 ps_product_shop ON (id_shop, active, visibility, id_product, id_category_default) 上有一个索引

SELECT 
      p.*, 
      ps.*, 
      stock.out_of_stock, 
      IFNULL(stock.quantity, 0) as quantity, 
      MAX(pas.id_product_attribute) id_product_attribute, 
      pas.minimal_quantity AS product_attribute_minimal_quantity, 
      pl.description, 
      pl.description_short, 
      pl.available_now,
      pl.available_later, 
      pl.link_rewrite, 
      pl.meta_description, 
      pl.meta_keywords, 
      pl.meta_title, 
      pl.name, 
      MAX(image_shop.id_image) id_image,
      il.legend, 
      m.name AS manufacturer_name, 
      cl.name AS category_default,
      DATEDIFF(ps.`date_add`, DATE_SUB(NOW(),INTERVAL 20 DAY)) > 0 AS new, 
      ps.price AS orderprice
   FROM 
      ( select @thisLanguage := 1 ) sqlvars,
      ps_product_shop ps
         INNER JOIN ps_product p
            ON ps.id_product = p.id_product

            INNER JOIN ps_category_product cp
               ON id_product = cp.id_product
               AND cp.id_category = 2 

            LEFT JOIN ps_product_attribute pa
               ON p.id_product = pa.id_product
               LEFT JOIN ps_product_attribute_shop pas
                  ON pa.id_product_attribute = pas.id_product_attribute
                  AND ps.id_shop = pas.id_shop
                  AND pas.default_on = 1

            LEFT JOIN ps_stock_available stock
               ON p.id_product = stock.id_product 
               AND ps.id_shop = stock.id_shop
               AND stock.id_shop_group = 0
               AND stock.id_product_attribute = IFNULL(pas.id_product_attribute, 0)

            LEFT JOIN ps_product_lang pl
               ON p.id_product = pl.id_product
               AND ps.id_shop = pl.id_shop 
               AND pl.id_lang = @thisLanguage

            LEFT JOIN ps_image i
               ON p.id_product = i.id_product
               LEFT JOIN ps_image_shop image_shop
                  ON i.id_image = image_shop.id_image 
                  AND ps.id_shop = image_shop.id_shop
                  AND image_shop.cover = 1
                  LEFT JOIN ps_image_lang il
                     ON image_shop.id_image = il.id_image
                     AND il.id_lang = @thisLanguage

            LEFT JOIN ps_manufacturer m
               ON p.id_manufacturer = m.id_manufacturer

         LEFT JOIN ps_category_lang cl
            ON ps.id_category_default = cl.id_category
            AND cl.id_shop = ps.id_shop
            AND cl.id_lang = @thisLanguage
   WHERE 
          ps.id_shop = 1
      AND ps.active = 1 
      AND ps.visibility IN ("both", "catalog") 
   GROUP BY 
      ps.id_product 
   ORDER BY 
      cp.position ASC
   LIMIT 
      0,8;

至于将多字段索引创建为单个索引，例如：

CREATE INDEX act_id_lang ON ps_category(active, id_category, id_lang, nleft, nright, level_depth);

给出一个简单的索引名称...在任何表上...使用所有键，因为它们将在查询中有效。

Answer 2

您应该在较长的使用时间后使用 mysqltuner。 这应该删除以下消息： MySQL 在过去 24 小时内启动 - 建议可能不准确