Inventory Feeds are currently created via API. The GraphQL API mutation to do this is createInventoryFeed. See GraphQL API schema for details of the required data for that mutation. 

When successful, the creation event will respond with a value to add as an S3 Bucket Policy. This value will allow the data to be replicated into your target bucket.

Required Data

InventorySourceInput

The

`source`

• type (InventoryDataType!): The type of inventory data. You can choose from:

  • `INVENTORY_CATALOGUE`

  • `INVENTORY_POSITION`

  • `VIRTUAL_CATALOGUE`

  • `VIRTUAL_POSITION`

• filters ([InventoryDataFilterInput]): Optional filters to apply to the inventory data. Each filter consists of:
  • name (InventoryFilterName!): The name of the filter. You can choose from:

    • `REF`

    • `TYPE`

    • `STATUS`

    • `CATALOGUE_REF`

  • operator (InventoryOperator!): The operator for the filter. You can choose from:

    • `IN`

    • `NOT_IN`

    • `EQUAL_TO`

    • `NOT_EQUAL_TO`

  • value (Json!): The value to filter by. This can be a single value or an array of values.

How Filters Work

When setting up filters, you first select an inventory data type and then optionally set filters based on that data type. Filters are combined using "AND" conditions, meaning all conditions must be met. Within each filter, "IN" and "NOT IN" operators can handle multiple values with an "OR" condition.

Examples

INVENTORY_POSITION Example

If you select

`INVENTORY_POSITION`

• STATUS   EQUAL_TO   "ACTIVE"
• CATALOGUE_REF IN  ["DEFAULT:1", "DEFAULT:67"]

This setup means:

• Fetch all inventory positions where the status is equal to "ACTIVE"
• AND the catalogue reference is either "DEFAULT:1" or "DEFAULT:67"

VIRTUAL_POSITION Example

If you select

`VIRTUAL_POSITION`

• STATUS   EQUAL_TO   "AT_RISK"
• CATALOGUE_REF IN  ["BASE:1", "AGGREGATE:1"]

This setup means:

• Fetch all virtual positions where the status is equal to "AT_RISK"
• AND the catalogue reference is either "BASE:1" or "AGGREGATE:1"

Example Mutation

1mutation {
2  createInventoryFeed(
3    input: {
4      ref: "Example_Inventory_Feed"
5      dataFormat: PARQUET
6      destination: {
7        type: AWS_S3
8        destinationAttributes: [
9          { key: AWS_ACCOUNT_ID, value: "AWS account ID" }
10          { key: AWS_S3_BUCKET_NAME, value: "Target S3 bucket name" }
11        ]
12      }
13      source: {
14        type: INVENTORY_POSITION
15        filters: [
16          {name: STATUS, operator: EQUAL_TO, value: "ACTIVE"}
17        ]
18      }
19    }
20  ) {
21    ref
22    dataFormat
23    destination {
24      type
25      destinationAttributes {
26        key
27        value
28      }
29    }
30    source  {
31      type
32      filters{
33        name
34        operator
35        value
36      }
37    }
38    attributes{
39      name
40      value
41    }
42  }
43}

Example Response

1{
2  "data": {
3    "createInventoryFeed": {
4      "ref": "Example_Inventory_Feed",
5      "dataFormat": "PARQUET",
6      "destination": {
7        "type": "AWS_S3",
8        "destinationAttributes": [
9          {
10            "key": "AWS_ACCOUNT_ID",
11            "value": "889803876844"
12          },
13          {
14            "key": "AWS account ID",
15            "value": "Target S3 bucket name"
16          }
17        ]
18      },
19      "source": {
20        "type": "INVENTORY_POSITION",
21        "filters": [
22          {
23            "name": "STATUS",
24            "operator": "EQUAL_TO",
25            "value": "ACTIVE"
26          }
27        ]
28      },
29      "attributes": [
30        {
31          "name": "roleArn",
32          "value": "arn:aws:iam::{source AWS account ID}:role/service-role/{IAM role name}"
33        }
34      ]
35    }
36  }
37}

You can add the new security policy to your target S3 bucket by following this guide from AWS: Adding a bucket policy by using the Amazon S3 console - Amazon Simple Storage Service.

Once this new policy has been applied, the data produced from the Inventory Feed will be replicated into the target S3 bucket as soon as there is a new piece of data produced by the Inventory feed.

Example Policy

1{
2  "Version": "2012-10-17",
3  "Id": "",
4  "Statement": [
5    {
6      "Sid": "Set permissions for objects",
7      "Effect": "Allow",
8      "Principal": {
9        "AWS": "arn:aws:iam::{SourceAccountId}:role/service-role/{RoleName}"
10      },
11      "Action": [
12        "s3:ReplicateObject",
13        "s3:ReplicateDelete"
14      ],
15      "Resource": "arn:aws:s3:::{DestinationBucketName}/*"
16    },
17    {
18      "Sid": "Set permissions on bucket",
19      "Effect": "Allow",
20      "Principal": {
21        "AWS": "arn:aws:iam::{SourceAccountId}:role/service-role/{RoleName}"
22      },
23      "Action": [
24        "s3:List*",
25        "s3:GetBucketVersioning",
26        "s3:PutBucketVersioning"
27      ],
28      "Resource": "arn:aws:s3:::{DestinationBucketName}"
29    }
30  ]
31}

After correctly creating the Inventory Feed and configuring the bucket policy on the target S3 bucket, you can update the Inventory Feed to “ACTIVE”. From this point, the Inventory Feed will begin running on the schedule as defined against the entity.

Example mutation

1mutation{
2    updateInventoryFeed(input: {
3  	    ref: "Example_Inventory_Feed",
4        status: "ACTIVE"
5    }) {
6        ref
7        status
8    }
9}

Example Mutation

1mutation {
2  createInventoryFeed(
3    input: {
4      ref: "Example_Inventory_Feed"
5      name: "Example Inventory Feed"
6      destination: {
7        type: AWS_S3
8        destinationAttributes: [
9          { key: AWS_ACCOUNT_ID, value: "AWS account ID" }
10          { key: AWS_S3_BUCKET_NAME, value: "Target S3 bucket name" }
11        ]
12      }
13      source: {
14        type: VIRTUAL_POSITION
15        filters: [
16          {name: STATUS, operator: EQUAL_TO, value: "ACTIVE"}
17        ]
18      }
19      dataFormat: PARQUET
20      frequencyCronExpression: "0 13 * * */2"
21    }
22  ) {
23    ref
24    name
25    destination {
26      type
27      destinationAttributes {
28        key
29        value
30      }
31    }
32    source  {
33      type
34      filters{
35        name
36        operator
37        value
38      }
39    }
40    attributes{
41      name
42      value
43    }
44    dataFormat
45    frequencyCronExpression
46  }
47}
48

Example mutation

1mutation {
2  updateInventoryFeed(
3    input: {
4      ref: "Example_Inventory_Feed"
5      name: "Test Inventory Feed"
6      frequencyCronExpression: "0 13 * * */2"
7    }
8  ) {
9    ref
10    name
11    frequencyCronExpression
12  }
13}
14

This query lets you pull data related to an individual Inventory Feed Run. To do this, you must know the unique reference of the Run you want to pull data from. 

This query lets you pull data about multiple Inventory Feed Runs at once. For example, if you’re looking for historical data from a specific Inventory Feed, this query will allow you to query every Run related to that feed.

To explore these queries and the other Inventory Feed queries, see our GraphQL API.

When updating an Inventory Feed to Active status, our system will validate that the target S3 bucket has the correct configuration.

Option 1: Confirm the target bucket has the correct policy applied

If the policy is correctly applied, check Option 2.

Option 2: Review the failure response

As part of the update mutation, the failure response will provide the context of what is failing. This can help you to fix the incorrect configuration.

An Inventory Feed Run may have failed for a range of reasons. Depending on these reasons, you may be unable to take proactive action.

Option 1: Check the Inventory Feed Run error message

The error message saved against an Inventory Feed Run will be able to provide context on what has caused the failure. If it is a configuration issue, you can update the Inventory Feed to fix this issue. Other times, it may be a random failure. At that point, the next Inventory Feed Run will catch the Inventory Feed back up to the latest view of Inventory Availability.

Option 2: Target S3 bucket configuration

If anything changes with the configuration of the target S3 bucket, this may cause Inventory Feed Runs to fail consistently. If this has happened, the first piece to try would be to reset the configuration. If that isn’t possible, you may need to create a new Inventory Feed to replace the broken one.