Requires pipeline agent v25.8.1 or later,
Core Operator v3.67.0 or later, and
Pipeline CLI v3.66.0 or later.
The S3 Input (One Time) source plugin
(name: go-s3-replay-plugin) lets you ingest data from an Amazon S3 bucket into
a telemetry pipeline.
This is a pull-based source plugin.
This plugin doesn’t support duplicates of itself within the same pipeline.
Supported telemetry types
This plugin for Chronosphere Telemetry Pipeline supports these telemetry types:
Requirements
To use the S3 Input (One Time) plugin, you must meet these requirements:
- Your IAM user or IAM role must have the
s3:GetObject
permission for your S3 bucket.
Configuration parameters
Use the parameters in this section to configure your plugin. The Telemetry Pipeline
web interface uses the values in the Name column to describe the parameters.
Items in the Key column are the YAML keys to use in
pipeline configuration files.
Required
| Name | Key | Description | Default |
|---|
| AWS Bucket Name | aws_bucket_name | Required. The name of the S3 bucket where your data is stored. | none |
| AWS Bucket Region | aws_bucket_region | Required. The name of the AWS region where your S3 bucket exists. For example: us-east1. | none |
| File Path | logs | Required. The glob format of the logs stored in your S3 bucket. For example: /logs/\*.log. | none |
| Line Buffer Max Size | max_line_buffer_size | The maximum line size the plugin will read from JSON or plain text files. | 10MiB |
Advanced
| Name | Key | Description | Default |
|---|
| S3 Assume Role ARN | assume_role_arn | The ARN of the role to assume in S3. This can be an ARN within the same account or across accounts. | none |
| S3 Role External ID | role_external_id | The external ID of the role to assume in S3. | none |
| S3 Endpoint | aws_s3_endpoint | The URL of your S3 endpoint. | none |
| S3 Read Concurrency | s3_read_concurrency | The maximum number of concurrent S3 GetObject calls that this plugin will make. | The number of logical CPUs allocated to each pipeline replica. |
| Memory Buffer Limit | mem_buf_limit | For pipelines with the Deployment or DaemonSet workload type only. Sets a limit for how much buffered data the plugin can write to memory, which affects backpressure. This value must follow Fluent Bit’s rules for unit sizes. If unspecified, no limit is enforced. | none |
Authentication methods
The S3 Input (One Time) plugin supports the following authentication methods:
EKS Pod Identities
To use EKS Pod Identities for authentication:
-
In AWS, configure EKS Pod Identities.
-
In Pipeline CLI, add the following flag to a
create pipeline
or update pipeline command:
calyptia {create|update} pipeline --service-account VALUE
VALUE with the name of the Kubernetes service account associated
with your Pods.
IMDS
To use IMDS for authentication:
IRSA
To use IRSA for authentication:
-
In AWS, set up IRSA for your EKS cluster.
-
Assign an IAM role to your Kubernetes service account.
-
In Pipeline CLI, add the following flag to a
create pipeline
or update pipeline command:
calyptia {create|update} pipeline --service-account VALUE
VALUE with the name of your Kubernetes service account.
Static credentials
To use static credentials for authentication:
- In Telemetry Pipeline, create create secrets
that contain the values of your IAM access keys.
These secrets must use the key names
AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY.
You don’t need to add an explicit reference
to these secrets in your pipeline configuration file. If secrets with the correct
key names are present, the S3 Input (One Time) plugin automatically detects these values
and uses them for authentication. Supported data types
The S3 Input (One Time) plugin can ingest JSON objects and plain text
from files stored in S3 buckets, including gzip-compressed files. Additionally,
this plugin can extract and ingest compressed and uncompressed files from
tar archives.
JSON
This plugin can ingest data from JSON files with these file extensions:
If a file contains only a single JSON object, this plugin creates
a new log from that object. If a file contains multiple
newline-delimited JSON (NDJSON)
objects, this plugin creates a new log from each JSON object within that file.
Key/value pairs from JSON objects are stored as key/value pairs in the resulting log.
For JSON files that use gzip compression (with file extensions such as .json.gzip
or .json.gz), this plugin decompresses each file before processing it accordingly.
Plain text
If a file doesn’t use a file extension that identifies it as a JSON file, the
S3 Input (One Time) plugin processes that file as plain text. It creates a new log from
each line of the file and stores the content in a key named _raw within the
resulting log.
For non-JSON files that use gzip compression (with file extensions that include
the .gzip or .gz suffix), this plugin decompresses each file before processing
it accordingly.
Tar archives
The plugin can extract and consume files from tar archives with these file extensions:
After the plugin extracts these files, it processes
any JSON and plain text data accordingly, but skips directories
and symbolic links.
If files inside a tar archive are gzip-compressed, this plugin decompresses those
files accordingly.
The S3 Input (One Time) plugin attaches the following metadata to each log:
__chrono_bucket: The name of the S3 bucket that contains the file from which
the log was created.__chrono_file: The key of the S3 object from which the log was created.__chrono_tar_file_entry: For data extracted from tar archives
only. The name of the tar archive that contained the file from which the log
was created.
Get started
To get started with the S3 Input (One Time) plugin, follow these steps.
-
Either create a new pipeline
or modify an existing pipeline.
-
For testing purposes, set the pipeline’s destination to
standard output.
-
Set the pipeline’s source to S3 Input (One Time), and then add values for all required
parameters, along with any optional parameters of your choosing.
-
Set up one of the supported authentication methods
for the S3 Input (One Time) source plugin.
-
In the Telemetry Pipeline web interface, go to the summary page for that pipeline.
-
In the Pipeline Output section,
click Get latest logs.
-
Review this log output to ensure that you’re receiving data from S3. If you don’t
receive any data, or if you encounter connection errors, review your plugin
configuration settings.
-
After you’ve confirmed that the S3 Input (One Time) plugin is functioning correctly,
you can overwrite the standard output destination with the destination where
you want to send your telemetry data.
