Question

我有一个汇总呼叫，如下所示。我相信发生角色查找时会造成大量延迟，因为有申请人，受雇人员和拒绝人员来从资源表中查找。资源集合中有10,000多个条目，因此查询大约需要6秒钟。我在做什么，这在这里是非常错误的？我看不到如何使用索引，因为所有查找都是通过默认已被索引的_id完成的。

应用程序，雇用字段和拒绝字段只是对象ID的数组，例如

applicants: [
 ObjectId('asldkajsdlkj'),
 ObjectId('asldkjaoksdjak')
]

任何帮助将不胜感激。在M0实例上花费了6秒，而在M10实例上并没有更快。

return db.collection('projects').aggregate([
  {
    $match: {
     agents: ObjectId('SOMETHING')
    }
  },
  {
    $lookup: {
      from: "agents",
      localField: "agents",
      foreignField: "_id",
      as: "agents"
    }
  },
  {
    $lookup: {
      from: "agencies",
      localField: "agency",
      foreignField: "_id",
      as: "agency"
    }
  }, 
  {
    $lookup: {
      from: "roles",
      let: { "roles": "$roles" },
      pipeline: [
        { $match: { $expr: { $in: [ "$_id", "$$roles" ] } } },
        {
          $lookup: {
            from:"resources",
            let: { "applicants": "$applicants" },
            pipeline: [
              { $match: { $expr: { $in: [ "$_id", "$$applicants" ] } } }
            ],
            as: "applicants"
          }
        },
        {
          $lookup: {
            from: "resources",
            let: { "hired": "$hired" },
            pipeline: [
              { $match: { $expr: { $in: [ "$_id", "$$hired" ] } } }
            ],
            as: "hired"
          }
        },
        {
          $lookup: {
            from: "resources",
            let: { "rejected": "$rejected" },
            pipeline: [
              { $match: { $expr: { $in: [ "$_id", "$$rejected" ] } } }
            ],
            as: "rejected"
          }
        },
        {
          $lookup: {
            from: "agents",
            localField: "hiring_agent",
            foreignField: "_id",
            as: "hiring_agent"
          }
        }
      ],
      as: "roles"
    }
  }
], {
  allowDiskUse: true
})

Answer 1

MongoDB旨在存储“煮熟”的数据。这意味着，在project集合中，您需要存储冗余信息以转换为所需的结果。

在您的情况下，由于to this error，$lookup集合的内部let - pipeline与role的性能下降。

尝试更改：

db.collection('projects').aggregate([
  {
    $match: {
     agents: ObjectId('SOMETHING')
    }
  },
  {
    $lookup: {
      from: "agents",
      localField: "agents",
      foreignField: "_id",
      as: "agents"
    }
  },
  {
    $lookup: {
      from: "agencies",
      localField: "agency",
      foreignField: "_id",
      as: "agency"
    }
  }, 
  {
    $lookup: {
      from: "roles",
      let: { "roles": "$roles" },
      pipeline: [
        { $match: { $expr: { $in: [ "$_id", "$$roles" ] } } },
        {
          $lookup: {
            from:"resources",
            localField: "applicants",
            foreignField: "_id",
            as: "applicants"
          }
        },
        {
          $lookup: {
            from: "resources",
            localField: "hired",
            foreignField: "_id",
            as: "hired"
          }
        },
        {
          $lookup: {
            from: "resources",
            localField: "rejected",
            foreignField: "_id",
            as: "rejected"
          }
        },
        {
          $lookup: {
            from: "agents",
            localField: "hiring_agent",
            foreignField: "_id",
            as: "hiring_agent"
          }
        }
      ],
      as: "roles"
    }
  }
], {
  allowDiskUse: true
})

管道中的多次查找

1 个答案: