Log objects store data about the execution of Pipelines and models. With this endpoint, you can check when your Pipelines and models were last executed and then trigger other processes in your analytics stack.
For example, if you use Kondado to create a feature table that feeds a machine learning model, you can check whether new data has been received that allows a new training cycle. Another example is when Kondado is used to populate a staging database that feeds a production database: you can query Pipeline logs to verify whether new data is available for consumption.
This endpoint uses a scrolling concept, where data is returned in batches of 50 records in ascending order. You must provide a parameter indicating the ID of the oldest record you want to retrieve.
Of course, you can also obtain log data directly using our free logs source, which consumes data from this API.
List Pipeline Logs
Description
Lists the logs of Pipelines that the user has access to. Only the last 90 days of data are available.
Method
GET
Endpoint
/pipeline_logs
URL Parameters
client_id
- Description: Parameter that indicates the account ID from which logs should be retrieved. If not provided, logs from all accounts the token has access to will be returned.
- Required: no
- Format: integer number
start_element_id
- Description: Parameter that indicates the ID of the oldest log record to be retrieved.
- Required: yes
- Format: integer number
Example Response
{
"success": true,
"data": {
"success": true,
"data": [
{
"id": 5181910,
"client_id": 1,
"pipeline_id": 702,
"execution_time_seconds": "0.486988",
"has_deltas": false,
"pipeline_name": "Kondado - Blog - wp_options",
"row_count": "256",
"createdAt": "2020-09-17T00:03:48.108Z",
"updatedAt": "2020-09-17T00:03:48.108Z",
"initial_savepoint": null,
"new_savepoint": null,
"external_ip": "3.220.239.218",
"is_billable": true,
"raw_row_count": "123",
"mb_estimate": "0.00231"
}
],
"pagination": {
"next_start_element_id": 5181910
}
}
}Response Description
The data object returns an array of JSON objects containing the 50 log records following the provided start_element_id.
The pagination object returns the next_start_element_id parameter, which can be used in a subsequent request to retrieve the next page of results. If it is null, there are no more log records after the provided start_element_id (end of pagination).
The JSON in the data object contains the following fields:
id
Log record ID
createdAt
Log creation timestamp (UTC) – indicates when the Pipeline finished executing.
updatedAt
Log update timestamp (UTC). Usually the same as createdAt, but may differ if the log is modified after creation.
pipeline_id
ID of the Pipeline to which the log belongs.
pipeline_name
Name of the Pipeline at the time the log was created.
client_id
ID of the client to which the Pipeline belongs.
has_deltas
Boolean indicating whether the Pipeline stored deltas when the log was created.
execution_time_seconds
Time in seconds the Pipeline took to execute for this log entry – not including the initial setup time (which involves tasks such as reading metadata about the Pipeline, source, and destination).
row_count
Number of records processed by the Pipeline.
is_billable
Boolean indicating whether the rows counted in row_count were billed as part of your plan or were free.
initial_savepoint
If the Pipeline is incremental, indicates the savepoint from which data was read. NULL for full Pipelines.
new_savepoint
If the Pipeline is incremental, indicates the new savepoint generated after execution. NULL for full Pipelines.
external_ip
IP address of the Kondado server that executed the Pipeline.
raw_row_count
Number of rows processed. This field is populated starting on June 12, 2025; for earlier dates, row_count is equivalent.
mb_estimate
Estimated size, in MB, of the data inserted by the execution.
List Model Logs
Description
Lists the logs of models that the user has access to.
Method
GET
Endpoint
/model_logs
URL Parameters
client_id
- Description: Parameter that indicates the account ID from which logs should be retrieved. If not provided, logs from all accounts the token has access to will be returned.
- Required: no
- Format: integer number
start_element_id
- Description: Parameter that indicates the ID of the oldest log record to be retrieved.
- Required: yes
- Format: integer number
Example Response
{
"success": true,
"data": {
"success": true,
"data": [
{
"id": 5219430,
"client_id": 1,
"model_id": 162,
"execution_time_seconds": 0.201923,
"model_name": "Kondado - Leads",
"row_count": 4915,
"createdAt": "2020-09-17T00:05:00.892Z",
"updatedAt": "2020-09-17T00:05:00.892Z"
}
],
"pagination": {
"next_start_element_id": 5219430
}
}
}Response Description
The data object returns an array of JSON objects containing the 50 log records following the provided start_element_id.
The pagination object returns the next_start_element_id parameter, which can be used in a subsequent request to retrieve the next page of results. If it is null, there are no more log records after the provided start_element_id (end of pagination).
The JSON in the data object contains the following fields:
id
Log record ID
createdAt
Log creation timestamp (UTC) – indicates when the model finished executing.
updatedAt
Log update timestamp (UTC). Usually the same as createdAt, but may differ if the log is modified after creation.
model_id
ID of the model to which the log belongs.
model_name
Name of the model at the time the log was created.
client_id
ID of the client to which the model belongs.
execution_time_seconds
Time in seconds the model took to execute for this log entry – not including the initial setup time (which involves tasks such as reading metadata about the Pipeline, source, and destination).
row_count
Number of rows present in the table generated by the model.
Retrieve and paginate execution logs via API
Use the Kondado API to fetch Pipeline and model execution logs with scrolling pagination, then integrate log data into your analytics workflows.
Authenticate and prepare your API request
Ensure your API token has access to the accounts whose logs you need. Decide whether to filter by client_id or retrieve logs across all accessible accounts.
Request the first batch of Pipeline logs
Send a GET request to /pipeline_logs with start_element_id set to 0 (or another starting ID). The endpoint returns 50 records in ascending order, including fields like execution_time_seconds, row_count, and has_deltas.
Paginate through results using next_start_element_id
Extract next_start_element_id from the pagination object. If it is not null, use it as the new start_element_id in your next request. Repeat until next_start_element_id is null, indicating the end of available logs.
Repeat the process for model logs
Apply the same scrolling pattern to the /model_logs endpoint. Model logs include model_id, model_name, and row_count for tables generated by data transformations.
Automate downstream processes based on log data
Use log timestamps and has_deltas flags to trigger events in your stack—for example, initiating a machine-learning training cycle when new data arrives, or confirming that a staging database is ready for production consumption.
Frequently asked questions
/pipeline_logs and /model_logs endpoints.pagination.next_start_element_id field in the response. If it is null, there are no more records after the current start_element_id and you have reached the end of the available logs.has_deltas boolean indicates whether the Pipeline stored deltas (incremental changes) when the log was created. This is useful for determining if new or changed data was processed during that execution.raw_row_count represents the number of rows processed, while row_count reflects the number of records processed by the Pipeline. The raw_row_count field is populated starting June 12, 2025; for earlier dates, row_count is equivalent. You can also use Kondado's platform features to monitor usage.