The Data Connector for Amazon S3 enables you to import the data from your JSON, TSV, and CSV files stored in an S3 bucket.

 

Prerequisites

• Basic knowledge of Arm Treasure Data

Use the TD Console to create your connection

You can use TD console to create your data connector.

Create a new connection

When you configure a data connection, you provide authentication to access the integration. In Treasure Data, you configure the authentication and then specify the source information.

Go to Catalog and search and select AWS S3. Click Create.

Using Credentials to Authenticate

You need a client ID and access keys to authenticate using credentials.

Register your credential. Set the following parameters.

• Endpoint: S3 endpoint login user name. You can find a region and endpoint information from AWS Document. (Ex. s3-ap-northeast-1.amazonaws.com)
• Authentication Method:
  • basic: uses access_key_id and secret_access_key to authenticate. See here.
    • Access Key ID
    • Secret access key
  • anonymous: uses anonymous access. This auth method can access only public files.
  • session: uses temporary-generated access_key_id, secret_access_key and session_token.
    • Access Key ID
    • Secret access key
    • Secret token

Click Continue. Name your new AWS S3 connection. Click Done.

Transfer data from AWS S3

Next, you need to create “New Transfer” on My Connections page. You can prepare an ad hoc DataConnector job or a scheduled data connector job. In this section, the following 4 steps are required.

Fetch from

You need to register the information that you would like to ingest.

• Bucket: S3 bucket name (Ex. your_bucket_name)
• Path Prefix: prefix of target keys. (Ex. logs/data_)
• Path Regex: regexp to match file paths. If a file path doesn’t match with this pattern, the file is skipped. (Ex. .csv$ # in this case, a file is skipped if its path doesn’t match with this pattern)
• Start after path: inserts last_path parameter so that first execution skips files before the path. (Ex. logs/data_20170101.csv)
• Incremental: enables incremental loading. If incremental loading is enabled, config diff for the next execution will include last_path parameter so that next execution skips files before the path. Otherwise, last_path will not be included.

Example: CloudFront

Amazon CloudFront is a web service that speeds up distribution of your static and dynamic web content. You can configure CloudFront to create log files that contain detailed information about every user request that CloudFront receives. If you enable logging, you can save CloudFront logfiles, shown as follows:

[your_bucket] - [logging] - [E231A697YXWD39.2017-04-23-15.a103fd5a.gz]
[your_bucket] - [logging] - [E231A697YXWD39.2017-04-23-15.b2aede4a.gz]
[your_bucket] - [logging] - [E231A697YXWD39.2017-04-23-16.594fa8e6.gz]
[your_bucket] - [logging] - [E231A697YXWD39.2017-04-23-16.d12f42f9.gz]

In this case, “Fetch from” setting should be as shown:

• Bucket: your_bucket
• Path Prefix: logging/
• Path Regex: .gz$ (Not Required)
• Start after path: logging/E231A697YXWD39.2017-04-23-15.b2aede4a.gz (Assuming that you want to import the logfiles from 2017-04-23-16.)
• Incremental: true (if you want to schedule this job.)

Preview

In this section, you can see a preview of data you configured. If you couldn’t reach this page, you might get any errors on Console. At that time, contact to support@treasuredata.com

If you would like to set specified column name, choose “Advanced Settings” button.

Advanced Settings

Advanced Settings allow you to edit guessed properties. Edit the following section, if you need to.

• Default timezone: changes Time zone of timestamp columns if the value itself doesn’t include time zone.
• Columns:
  • Name: changes a name of the column. Column name is supported consisting of lowercase alphabets, numbers, and “_” only.
  • Type: parses a value as a specified type. And then, it stores after converting to TreasureData schema.
    • boolean
    • long
    • timestamp: will be imported as String type at TreasureData (Ex. 2017-04-01 00:00:00.000)
    • double
    • string
    • json
• Total file count limit: maximum number of files to read. (optional)

Transfer to

In this phase, select your target database/table you want to import to.

• Mode: Append/Replace
• Partition key Seed: choose the long or timestamp column as the partitioning time. As default time column, it’s used upload_time with using add_time filter.

When

In this phase, you can set an ad hoc or schedule configuration for your job.

• When
  • Once now: set one time job.
  • Repeat…
    • Schedule: accepts these three options: @hourly, @daily and @monthly and custom cron.
    • Delay Transfer: add a delay of execution time.
  • TimeZone: supports extended timezone formats like ‘Asia/Tokyo’.

My Input Transfers

Finally, your DataConnector jobs are listed up on My Input Transfer. In this page, you can edit your existing job. Also, you can see a DataConnector job’s detail, which run before, by clicking Last Transfer section.

Scheduled Execution

You can schedule periodic data connector execution for incremental S3 file import. We configure our scheduler carefully to ensure high availability. By using this feature, you no longer need a cron daemon on your local datacenter.

For the scheduled import, the Data Connector for Amazon S3 imports all files that match with the specified prefix (e.g. path_prefix: path/to/sample_ –> path/to/sample_201501.csv.gz, path/to/sample_201502.csv.gz, …, path/to/sample_201505.csv.gz) initially and remembers the last path (path/to/sample_201505.csv.gz) for the next execution.

On the second and on subsequent runs, the connector imports only files that comes after the last path in alphabetical (lexicographic) order. (path/to/sample_201506.csv.gz, …)

Create Schedule

A new schedule can be created using the td connector:create command. The following are required: the name of the schedule, the cron-style schedule, the database and table where their data will be stored, and the Data Connector configuration file.

$ td connector:create daily_import "10 0 * * *" \
    td_sample_db td_sample_table load.yml

It’s also recommended to specify the --time-column option, because Treasure Data’s storage is partitioned by time (see also data partitioning)

$ td connector:create daily_import "10 0 * * *" \
    td_sample_db td_sample_table load.yml \
    --time-column created_at

List All Schedules

You can see the list of currently scheduled entries by running the command td connector:list.

$ td connector:list
+--------------+--------------+----------+-------+--------------+-----------------+------------------------------------------+
| Name         | Cron         | Timezone | Delay | Database     | Table           | Config                                   |
+--------------+--------------+----------+-------+--------------+-----------------+------------------------------------------+
| daily_import | 10 0 * * *   | UTC      | 0     | td_sample_db | td_sample_table | {"in"=>{"type"=>"s3", "access_key_id"... |
+--------------+--------------+----------+-------+--------------+-----------------+------------------------------------------+

Show Schedule Settings And History

td connector:show shows the execution setting of a schedule entry.

% td connector:show daily_import
Name     : daily_import
Cron     : 10 0 * * *
Timezone : UTC
Delay    : 0
Database : td_sample_db
Table    : td_sample_table
Config
---
in:
  type: s3
  access_key_id: XXXXXXXXXX
  secret_access_key: YYYYYYYYYY
  endpoint: s3.amazonaws.com
  bucket: sample_bucket
  path_prefix: path/to/sample_
  parser:
    charset: UTF-8
    ...

td connector:history shows the execution history of a schedule entry. To investigate the results of each individual run, use td job <jobid>.

% td connector:history daily_import
+--------+---------+---------+--------------+-----------------+----------+---------------------------+----------+
| JobID  | Status  | Records | Database     | Table           | Priority | Started                   | Duration |
+--------+---------+---------+--------------+-----------------+----------+---------------------------+----------+
| 578066 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-18 00:10:05 +0000 | 160      |
| 577968 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-17 00:10:07 +0000 | 161      |
| 577914 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-16 00:10:03 +0000 | 152      |
| 577872 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-15 00:10:04 +0000 | 163      |
| 577810 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-14 00:10:04 +0000 | 164      |
| 577766 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-13 00:10:04 +0000 | 155      |
| 577710 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-12 00:10:05 +0000 | 156      |
| 577610 | success | 10000   | td_sample_db | td_sample_table | 0        | 2015-04-11 00:10:04 +0000 | 157      |
+--------+---------+---------+--------------+-----------------+----------+---------------------------+----------+
8 rows in set

Delete Schedule

td connector:delete will remove the schedule.

$ td connector:delete daily_import

FAQ for the S3 Data Connector

Q: Data Connector for S3 job is running for a long time, what can I do?

• Check the count of S3 files that your connector job is ingesting. If there are over 10,000 files, the performance degrades. To mitigate this issue, you can:
  • Narrow path_prefix option and reduce the count of S3 files.
  • Set 268,435,456 (256MB) to min_task_size option.

Q: How do I troubleshoot data import problems?

Review the job log. Warning and errors provide information about the success of your import. For example, you can identify the source file names associated with import errors.

Appendix

A) Optional Alternative: Use the CLI to Configure the Connector

Before setting up the connector, install the ‘td’ command. Install the most current Treasure Data Toolbelt.

Step 1: Create Seed Config File (seed.yml)

First, prepare the seed.yml as shown in the following example, with your AWS access key and secret access key. You must also specify bucket name, and source file name (or prefix for multiple files).

in:
  type: s3
  access_key_id: XXXXXXXXXX
  secret_access_key: YYYYYYYYYY
  bucket: sample_bucket
  # path to the *.json or *.csv or *.tsv file on your s3 bucket
  path_prefix: path/to/sample_file 
out:
  mode: append

The Data Connector for Amazon S3 imports all files that match the specified prefix. (e.g. path_prefix: path/to/sample_ –> path/to/sample_201501.csv.gz, path/to/sample_201502.csv.gz, …, path/to/sample_201505.csv.gz)

For more details on available out modes, see Appendix.

Step 2: Guess Fields (Generate load.yml)

Second, use connector:guess. This command automatically reads the source files, and assesses (uses logic to guess) the file format and its field/columns.

$ td connector:guess seed.yml -o load.yml

If you open up load.yml, you’ll see the assessed file format definitions including file formats, encodings, column names, and types.

in:
  type: s3
  access_key_id: XXXXXXXXXX
  secret_access_key: YYYYYYYYYY
  bucket: sample_bucket
  path_prefix: path/to/sample_file
  parser:
    charset: UTF-8
    newline: CRLF
    type: csv
    delimiter: ','
    quote: '"'
    escape: ''
    skip_header_lines: 1
    columns:
    - name: id
      type: long
    - name: company
      type: string
    - name: customer
      type: string
    - name: created_at
      type: timestamp
      format: '%Y-%m-%d %H:%M:%S'
out:
  mode: append

Then, you can see a preview of the data using the td connector:preview command.

$ td connector:preview load.yml
+-------+---------+----------+---------------------+
| id    | company | customer | created_at          |
+-------+---------+----------+---------------------+
| 11200 | AA Inc. |    David | 2015-03-31 06:12:37 |
| 20313 | BB Imc. |      Tom | 2015-04-01 01:00:07 |
| 32132 | CC Inc. | Fernando | 2015-04-01 10:33:41 |
| 40133 | DD Inc. |    Cesar | 2015-04-02 05:12:32 |
| 93133 | EE Inc. |     Jake | 2015-04-02 14:11:13 |
+-------+---------+----------+---------------------+

If the system detects your column name or column type unexpectedly, modify load.yml directly and preview again.

Currently, the Data Connector supports parsing of “boolean”, “long”, “double”, “string”, and “timestamp” types.

Step 3: Execute Load Job

Finally, submit the load job. It may take a couple of hours depending on the size of the data. Specify the Treasure Data database and table where the data should be stored.

It’s also recommended to specify --time-column option, because Treasure Data’s storage is partitioned by time (see data partitioning) If the option is not provided, the data connector will choose the first long or timestamp column as the partitioning time. The type of the column specified by --time-column must be either of long and timestamp type.

If your data doesn’t have a time column you can add a time column by using add_time filter option. For more details see add_time filter plugin

$ td connector:issue load.yml --database td_sample_db --table td_sample_table \
  --time-column created_at

The connector:issue command assumes that you have already created a database(td_sample_db) and a table(td_sample_table). If the database or the table do not exist in TD, this command will not succeed, so create the database and table manually or use --auto-create-table option with td connector:issue command to auto create the database and table:

$ td connector:issue load.yml --database td_sample_db --table td_sample_table --time-column created_at --auto-create-table

If you have a field called time, you don’t have to specify the --time-column option.

$ td connector:issue load.yml --database td_sample_db --table td_sample_table

B) Modes for out plugin

You can specify file import mode in out section of seed.yml.

append (default)

This is the default mode and records are appended to the target table.

in:
  ...
out:
  mode: append

replace

This mode replaces data in the target table. Any manual schema changes made to the target table remain intact with this mode.

in:
  ...
out:
  mode: replace

C) IAM Permissions

The IAM credentials specified in the YML configuration file and used for the connector:guess and connector:issue commands need to be allowed permissions for the AWS S3 resources that they need to access. If the IAM user does not possess these permissions, configure the user with one of the predefined Policy Definitions or create a new Policy Definition in JSON format.

The following example is based off the Policy Definition reference format, giving the IAM user read only (through GetObject and ListBucket actions) permission for the your-bucket bucket:

{
  "Version": "2015-11-11",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:GetObject",
        "s3:ListBucket"
      ],
      "Resource": [
        "arn:aws:s3:::your-bucket",
        "arn:aws:s3:::your-bucket/*"
      ]
    }
  ]
}

Replace your-bucket with the actual name of your bucket.

D) Use AWS Security Token Service (STS) as temporary credentials provider

In certain cases, IAM basic authentication through access_key_id and secret_access_key might be too risky (although the secret_access_key is never clearly shown when a job is executed or after a session is created).

The S3 data connector can use AWS Secure Token Service (STS) provided Temporary Security Credentials. Using AWS STS, any IAM user can use his own access_key_id and secret_access_key to create a set of temporary new_access_key_id, new_secret_access_key, and session_token keys with an associated expiration time, after which the credentials become invalid.
There are essentially 2 types of Temporary Security Credentials:

1. Session Token
   The simplest Security Credentials with an associated expiration time. The temporary credentials give access to all resources the original IAM credentials used to generate them had. These credentials are valid as long as they are not expired and the permissions of the original IAM credentials don’t change.
2. Federation Token
   Adds an extra layer of permission control over the Session Token above. When generating a Federation Token, the IAM user is required to specify a Permission Policy definition. The scope can be used to further narrow down which of the resources, accessible to the IAM user, the bearer of the Federation Token should get access to. Any Permission Policy definition can be used but the scope of the permission is limited to only all or a subset of the permissions the IAM user used to generate the token had. As for the Session Token, the Federation Token credentials are valid as long as they are not expired and the permissions associated to the original IAM credentials don’t change.

AWS STS Temporary Security Credentials can be generated using the AWS CLI or the AWS SDK in the language of your choice.

Session Token

$ aws sts get-session-token --duration-seconds 900
{
    "Credentials": {
        "SecretAccessKey": "YYYYYYYYYY",
        "SessionToken": "ZZZZZZZZZZ",
        "Expiration": "2015-12-23T05:11:14Z",
        "AccessKeyId": "XXXXXXXXXX"
    }
}

Federation Token

$ aws sts get-federation-token --name temp_creds --duration-seconds 900 \
  --policy '{"Statement": [{"Effect": "Allow", "Action": ["s3:GetObject", "s3:ListBucket"], "Resource": "arn:aws:s3:::bucketname"}]}'
{
    "FederatedUser": {
        "FederatedUserId": "523683666290:temp_creds",
        "Arn": "arn:aws:sts::523683666290:federated-user/temp_creds"
    },
    "Credentials": {
        "SecretAccessKey": "YYYYYYYYYY",
        "SessionToken": "ZZZZZZZZZZ",
        "Expiration": "2015-12-23T06:06:17Z",
        "AccessKeyId": "XXXXXXXXXX"
    },
    "PackedPolicySize": 16
}

where: * temp_cred is the name of the Federated token/user * bucketname is the name of the bucket to give access to. Refer to the ARN specification for more details * s3:GetObject and s3:ListBucket are the basic read operation for a AWS S3 bucket.

When your Temporary Security Credentials are generated, copy the SecretAccessKey, AccessKeyId, and SessionToken in your seed.yml file as follows.

in:
  type: s3
  auth_method: session
  access_key_id: XXXXXXXXXX
  secret_access_key: YYYYYYYYYY
  session_token: ZZZZZZZZZZ
  bucket: sample_bucket
  path_prefix: path/to/sample_file

and execute the Data Connector for S3 as usual.