private-message-topic-tracking-state.json 加载缓慢

自 2.8.0.beta5 版本起,加载群组私信收件箱需要 4 秒以上,其中大部分时间(2-3 秒)消耗在 private-message-topic-tracking-state.json 的大型 SQL 查询上。

我们约有 80 万个主题、380 万用户,并且可能存在一个庞大的未读私信与用户关联图,但在查询分析中我并未发现太多值得注意的问题(可能是我不太擅长解读)。

在 beta5 之前一切正常,现在私信收件箱变得非常难以追踪,令人烦恼。

1 个赞

嗯,我们在 Meta 上的平均耗时约为 200 毫秒。

不过,现在访问个人资料中的消息会发起 3 次独立的请求,这确实有点可疑。而且你提到的那个请求返回了一个包含 900 个对象的 92kB JSON 文件,这让我很震惊 :scream:

2 个赞

这似乎确实会在中途使 miniprofiler 的状态失效,导致我甚至无法获取 SQL 查询以正确地进行 EXPLAIN 分析。

2 个赞

是的,那条路由转换表现不太正常。

顺便提一下,我提取了一个执行计划:ieWq | explain.depesz.com

3 个赞

此修复应能将其降至约 13 毫秒。

请注意,未读消息越多,查询的开销就越大。

@sam 我还需要修复另一件事:标记未读为已读时,不会更新我们为普通主题和私信主题设置的 first_unread_at 优化。

4 个赞

在 cherry-picking 此更改并重启 unicorn 后的 1.73 秒,但我和我的组的“未读时间”可能滞后得相当多(我当前邮箱仅显示 59 条新消息和 4 条未读——这个路由是按用户划分的吗?不过即使对于“所有消息”,我也只有 61 条新消息和 7 条未读)。

1 个赞

未读时间是否与您的收件箱中的未读消息相关联?

1 个赞
discourse=# select first_unread_pm_at from group_users where user_id = 2334 order by first_unread_pm_at asc;
     first_unread_pm_at
----------------------------
 2021-09-03 06:54:13.493514
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
 2021-09-03 07:05:33.138986
(17 rows)
discourse=# select first_unread_pm_at from user_stats where user_id = 2334;
     first_unread_pm_at
----------------------------
 2021-09-03 07:05:00.565522
(1 row)

奇怪的是,这个时间范围甚至不算特别大,所以应该是其他条件导致了性能缓慢(不过,神秘的是,在忽略几条未读消息后,查询速度确实提升到了 700 毫秒,这还可以接受。我不确定我之前的 first_unread_pm_at 范围是多少,也许是几周?)。

有没有办法让我在消息页面刷新并清除所有旧请求后,从 miniprofiler 中获取查询语句,以便我运行 EXPLAIN?

这确实很奇怪,但我无法在开发环境中复现此问题。下周我会花时间进一步调查。

1 个赞

已在以下链接中修复:

1 个赞

谢谢!我会尽量记住在下一次重建/重启窗口时再次检查。

1 个赞

拿到了查询计划/EXPLAIN ANALYZE 结果(最高达 1.3 秒:confused):CvUX | explain.depesz.com

在执行 reindex index concurrently index_topics_on_timestamps_private; 后,速度有所提升 (768ms)。这可能是因为之前大量的私信(DM)记录所致(在清理之前,叙事机器人严重污染了私信表)……不过在该索引上的查询速度仍然较慢。

你能在 Rails 控制台中为我运行 Topic.where(archetype: Archetype.private_message).count 吗?

顺便提一下,该查询现在是异步执行的,因此不会阻塞消息页面或私信的加载。

[1] pry(main)> Topic.where(archetype: Archetype.private_message).count
=> 543855

情况可能更糟,我们最近已经清理了所有遗留的叙事机器人主题。:sweat_smile: 不过还是比我预期的要多!

啊,难怪那个查询要过一会儿才在 miniprofiler 中显示。这很好!

我们最近再次更新了以获取此功能 https://github.com/discourse/discourse/pull/13603(即当前版本为 2.8.0beta6)。

看起来“获取用户所有私信”的功能,最终会指向 FEATURE: New and Unread messages for user personal messages. by tgxworld · Pull Request #13603 · discourse/discourse · GitHub

SELECT "topics"."id", "topics"."title", "topics"."last_posted_at", "topics"."created_at", "topics"."updated_at", "topics"."views", "topics"."posts_count", "topics"."user_id", "topics"."last_post_user_id", "topics"."reply_count", "topics"."featured_user1_id", "topics"."featured_user2_id", "topics"."featured_user3_id", "topics"."deleted_at", "topics"."highest_post_number", "topics"."like_count", "topics"."incoming_link_count", "topics"."category_id", "topics"."visible", "topics"."moderator_posts_count", "topics"."closed", "topics"."archived", "topics"."bumped_at", "topics"."has_summary", "topics"."archetype", "topics"."featured_user4_id", "topics"."notify_moderators_count", "topics"."spam_count", "topics"."pinned_at", "topics"."score", "topics"."percent_rank", "topics"."subtype", "topics"."slug", "topics"."deleted_by_id", "topics"."participant_count", "topics"."word_count", "topics"."excerpt", "topics"."pinned_globally", "topics"."pinned_until", "topics"."fancy_title", "topics"."highest_staff_post_number", "topics"."featured_link", "topics"."reviewable_score", "topics"."image_upload_id", "topics"."slow_mode_seconds" FROM "topics" LEFT OUTER JOIN topic_users AS tu ON (topics.id = tu.topic_id AND tu.user_id = 1234) LEFT JOIN group_archived_messages gm ON gm.topic_id = topics.id
LEFT JOIN user_archived_messages um
ON um.user_id = 1234
AND um.topic_id = topics.id WHERE "topics"."deleted_at" IS NULL AND (topics.id IN (
SELECT topic_id
FROM topic_allowed_users
WHERE user_id = 1234
 UNION ALL
SELECT topic_id FROM topic_allowed_groups
 WHERE group_id IN (
SELECT group_id FROM group_users WHERE user_id = 1234
  )
)) AND "topics"."archetype" = 'private_message' AND "topics"."visible" = TRUE AND (um.user_id IS NULL AND gm.topic_id IS NULL) ORDER BY topics.bumped_at DESC LIMIT 30;

该查询基本导致我们的 RDS 服务器挂起。我们团队的一位成员建议,在连接之前先利用现有索引过滤数据。优化后的查询运行时间小于 100 毫秒。

SELECT "topics_filter"."id", "topics_filter"."title", "topics_filter"."last_posted_at", "topics_filter"."created_at", "topics_filter"."updated_at",
"topics_filter"."views", "topics_filter"."posts_count", "topics_filter"."user_id", "topics_filter"."last_post_user_id", "topics_filter"."reply_count", "topics_filter"."featured_user1_id",
"topics_filter"."featured_user2_id", "topics_filter"."featured_user3_id", "topics_filter"."deleted_at", "topics_filter"."highest_post_number", "topics_filter"."like_count",
"topics_filter"."incoming_link_count", "topics_filter"."category_id", "topics_filter"."visible", "topics_filter"."moderator_posts_count", "topics_filter"."closed",
"topics_filter"."archived", "topics_filter"."bumped_at", "topics_filter"."has_summary", "topics_filter"."archetype", "topics_filter"."featured_user4_id", "topics_filter"."notify_moderators_count",
"topics_filter"."spam_count", "topics_filter"."pinned_at", "topics_filter"."score", "topics_filter"."percent_rank", "topics_filter"."subtype", "topics_filter"."slug",
"topics_filter"."deleted_by_id", "topics_filter"."participant_count", "topics_filter"."word_count", "topics_filter"."excerpt", "topics_filter"."pinned_globally",
"topics_filter"."pinned_until", "topics_filter"."fancy_title", "topics_filter"."highest_staff_post_number", "topics_filter"."featured_link", "topics_filter"."reviewable_score",
"topics_filter"."image_upload_id", "topics_filter"."slow_mode_seconds"
FROM (select * from "topics"
      where "topics"."deleted_at" IS NULL
      and (topics.id IN ( SELECT topic_id FROM topic_allowed_users WHERE user_id = 1234
                          UNION ALL
                          SELECT topic_id FROM topic_allowed_groups  WHERE group_id IN ( SELECT group_id FROM group_users WHERE user_id = 1234 )
                         ))
      AND "topics"."archetype" = 'private_message'
      AND "topics"."visible" = TRUE
      )   as topics_filter
LEFT OUTER JOIN (select topic_id from topic_users where user_id = 1234) AS tu ON (topics_filter.id = tu.topic_id )
LEFT JOIN group_archived_messages gm ON gm.topic_id = topics_filter.id
LEFT JOIN ( select user_id, topic_id from user_archived_messages where user_id = 1234) um
ON um.topic_id = topics_filter.id
WHERE (um.user_id IS NULL AND gm.topic_id IS NULL) ORDER BY topics_filter.bumped_at DESC LIMIT 30;

供参考:topics 表超过 1000 万行,topic_users 表超过 5000 万行。

是否有可能重构该查询以某种方式解决这些问题?

编辑:

澄清一下,当我们检查运行时间超过 2 分钟甚至 5 分钟的长查询时,出现的只有这些查询。在应用最新更改后,我们的 Postgres CPU 持续处于 99% 的使用率。当我们终止这些查询时,CPU 使用率会恢复正常。

3 个赞

@tgxworld 很抱歉直接提及您,但既然您是作者,我想请教您是否有任何建议或反馈。

1 个赞

据我所知,@sam 一直将其描述为一项高风险且危险的功能。如果这里出现重大回归,我并不会感到意外。

我们在发布此内容后发现了这个关于 bug 的讨论帖:Slow loading on private-message-topic-tracking-state.json - #9 by tgxworld

看起来在标记为 v2.8.0beta6 的提交之后,可能还有一个提交似乎解决了部分性能问题。我们打算尝试针对该特定提交引用重新构建,看看是否能缓解任何问题。

2 个赞

您当前的 RDS 实例类型和数据库总大小是多少?

1 个赞

db.m5.2xlarge

磁盘使用量约为 250GB。

1 个赞