Optimize your agent

This is the official documentation of the @forestadmin/agent Node.js agent.

Your new agent is up and running, congratulations!

Because of the internal changes, you might already have noticed a performance improvement when you migrated your agent. Let's see how you can optimize your agent even more.

Computed fields

Replace queries with the dependencies option

A common pattern in the old agent was to make queries in the get function of a computed field to fetch relations.

This is now natively supported with the dependencies option, and it is much faster.

If you ported your agent by using the simpler route, which is copying the old agent code to the new agent, you can probably replace a lot of your queries with the dependencies option.

Before

agent.customizeCollection('post', postCollection => {
  postCollection.addField('authorFullName', {
    columnType: 'String',
    dependencies: ['authorId'],
    getValues: posts =>
      posts.map(async post => {
        // Those async queries take a long time and are performed in parallel with
        // the other queries
        const author = await models.authors.findOne({
          where: { id: post.authorId },
        });

        return `${author.firstName} ${author.lastName}`;
      }),
  });
});

After

agent.customizeCollection('post', postCollection => {
  postCollection.addField('authorFullName', {
    columnType: 'String',
    // The agent will automatically fetch the author collection and join it with the
    // post collection by performing a INNER JOIN on the authorId field.
    // This is _much_ faster than performing a query for each post

    dependencies: ['author:firstName', 'author:lastName'],
    getValues: posts =>
      posts.map(post => `${post.author.firstName} ${post.author.lastName}`),
  });
});

or (better)

// Define the field on the author collection
agent.customizeCollection('author', authorCollection => {
  authorCollection.addField('fullName', {
    columnType: 'String',
    dependencies: ['firstName', 'lastName'],
    getValues: authors =>
      authors.map(author => `${author.firstName} ${author.lastName}`),
  });
});

// And then import it on the post collection
agent.customizeCollection('post', postCollection => {
  postCollection.importField('authorFullName', { path: 'author:fullName' });
});

Move async calls outside of the hot loop

The new computed field syntax works in batch mode.

If you have computed fields that take a long time to compute, you may want to use this new syntax to optimize them.

Before

agent.customizeCollection('users', users => {
  users.addField('full_address', {
    columnType: 'String',
    dependencies: ['address_id'],
    getValues: users =>
      users.map(async user => {
        // Get the address for each user individually
        // This is very slow if you are displaying a lot of users in a table
        const address = await geoWebService.getAddress(customer.address_id);

        return [address.line_1, address.line_2, address.city, address.country].join(
          '\n',
        );
      }),
  });
});

After

agent.customizeCollection('users', users => {
  users.addField('full_address', {
    columnType: 'String',
    dependencies: ['address_id'],
    getValues: async users => {
      // Get all the addresses in a single request
      // This is much faster if you are displaying a lot of users in a table
      const addresses = await geoWebService.getAddresses(
        users.map(user => user.address_id),
      );

      // Return the full address for each user
      return users.map(user => {
        const addr = addresses.find(addr => addr.id === user.address_id);
        return [addr.line1, addr.line2, addr.city, addr.country].join('\n');
      });
    },
  });
});

Avoid duplicate queries

If you happen to have computed fields that are similar, or that depend on the same data, you can make fields that depend on other fields.

Before

agent.customizeCollection('users', users => {
  users.addField('firstName', {
    columnType: 'String',
    dependencies: ['id'],
    getValues: async users => {
      const userIds = users.map(user => user.id);
      const userInfos = await authenticationWebService.getUserInfo(userIds);

      return users.map(
        user => userInfos.find(userInfo => userInfo.id === user.id).firstName,
      );
    },
  });

  users.addField('lastName', {
    columnType: 'String',
    dependencies: ['id'],
    getValues: async users => {
      const userIds = users.map(user => user.id);
      const userInfos = await authenticationWebService.getUserInfo(userIds);

      return users.map(
        user => userInfos.find(userInfo => userInfo.id === user.id).lastName,
      );
    },
  });
});

After

Install the @forestadmin/plugin-flattener plugin.

npm install @forestadmin/plugin-flattener

Fetch all the user info in a single request, and then flatten the result.

const { flattenColumn } = require('@forestadmin/plugin-flattener');

agent.customizeCollection('users', users => {
  users.addField('userInfo', {
    columnType: { firstName: 'String', lastName: 'String' },
    dependencies: ['id'],
    getValues: async users => {
      const userIds = users.map(user => user.id);
      const userInfos = await authenticationWebService.getUserInfo(userIds);

      return users.map(user => userInfo.find(userInfo => userInfo.id === user.id));
    },
  });

  users.use(flattenColumn, { columnName: 'userInfo' });
});

Search

The new agent introduced the capability to customize the search behavior of your agent depending on the search query.

This is a very powerful feature that allows you to make sure that your agent is fast.

Documentation about this feature is available here.