ZenotiHelp

Data freshness and refresh cycles

ZTZenoti Docs Team Updated October 27, 2026 7 min readGuide

Analytics — Redshift Data Share delivers data through automated refresh cycles that keep your data warehouse current with your Zenoti operational system. This article explains how data refreshes work, how often they run, and how you can verify data freshness.


On this page


How often data refreshes

Incremental data refreshes run approximately every 1--2 hours during normal operation. Exact timing varies depending on the volume of data being processed and overall system conditions. Do not rely on an exact interval for downstream pipeline scheduling.

Each refresh cycle identifies records that have changed in the Zenoti operational system since the last cycle, processes those changes, and updates the Redshift data warehouse. After a refresh completes, any SQL query you run against the Redshift endpoint reflects the updated data.

Note: Data in the Redshift data warehouse reflects the state as of the last completed refresh cycle. There is an inherent lag between when a transaction occurs in Zenoti and when it appears in the data warehouse.

Full load vs incremental load

There are two types of data loads:

Load type When it runs What it contains
Full load Once, during initial onboarding All historical data from your Zenoti organization
Incremental load Every 1--2 hours after onboarding Only records created, updated, or deleted since the previous cycle

The full load gives you a complete historical baseline. Every subsequent incremental load applies only the changes from that cycle, keeping the data warehouse current without reprocessing your entire dataset.

How change detection works

The system tracks changes at the record level. During each incremental cycle, only records that have been created, updated, or deleted since the previous cycle are extracted and processed. This change-detection approach keeps refresh cycles efficient -- even organizations with large data volumes receive timely updates because only the delta is processed.

The data pipeline

Data flows from your Zenoti operational system through automated processing stages to the Redshift data warehouse. You do not need to manage, monitor, or interact with any part of this pipeline -- Zenoti handles all processing automatically.

Conceptually, the pipeline operates in four stages:

Stage 1 -- Extraction

Changed records are extracted from the Zenoti operational system. The extraction process identifies what has been created, updated, or deleted since the last refresh cycle. For new customer or center onboarding, this stage extracts all historical data to perform the initial full load.

Stage 2 -- Transformation

Business logic is applied to structure the extracted data into the star schema format. This stage creates and updates the dimension tables (descriptive attributes) and fact tables (measurable events).

Data sources vary by table type:

  • Some tables are derived from business transactions that require transformation logic (such as invoice items and appointment records)
  • Other tables are sourced directly from operational reference data (such as services, products, and center configurations)

Stage 3 -- Loading

Transformed data is loaded into the Redshift data warehouse. During this stage:

  • Dimension tables are updated with current values (update-insert pattern)
  • Fact tables are refreshed using a delete-insert pattern to maintain data consistency
  • Consolidated tables are rebuilt by joining the updated dimension and fact tables

Stage 4 -- Export

For customers who also subscribe to Analytics — Amazon S3, the same data is exported as CSV files to their S3 bucket. This stage runs after the Redshift data warehouse is updated, so both delivery mechanisms reflect the same data.

Processing modes for reliability

Zenoti maintains redundant processing paths to ensure data warehouse reliability. Two operating modes exist:

  • Regular mode -- the standard processing path used during normal operation
  • Failover mode -- an alternative processing path that activates automatically if the primary path encounters issues

Both modes produce identical output in the Redshift data warehouse. As a customer, you do not need to know which mode is active -- your queries return the same data regardless. The redundancy ensures that refresh cycles continue even if one processing path is temporarily unavailable.

Consolidated table refresh

Consolidated tables are rebuilt during each refresh cycle by joining the latest dimension and fact data. The rebuild process works differently depending on the table size:

  • Large consolidated tables (such as sales, collections, and appointments) are processed in batches. The system processes a configurable number of records per batch, allowing large tables to be refreshed incrementally without consuming excessive resources.
  • Small and custom consolidated tables are fully rebuilt each cycle. Because they contain fewer records, a complete rebuild is efficient and ensures full consistency.

To minimize query disruption, consolidated tables use an atomic swap approach: the new version of a table is built in a staging area, then swapped into place as the live table. This means your queries never see a partially rebuilt consolidated table -- you always query either the previous complete version or the new complete version.

S3 Data Share as an alternative delivery

If your analytics workflow requires file-based data delivery instead of (or in addition to) direct SQL access, Zenoti also offers Analytics — Amazon S3. S3 Data Share exports the same data warehouse contents as CSV files to an Amazon S3 bucket.

Both products share the same underlying data, table schemas, and refresh cadence. You can use one or both depending on your needs.

Tracking data freshness with Bi_DimETLRefreshLog

The Bi_DimETLRefreshLog table records when data was last refreshed for each center in your organization. Query this table to determine the most recent successful refresh for any given center.

SELECT
    c.centername,
    MAX(r.centerrefreshdatetime) AS last_refresh
FROM Bi_DimETLRefreshLog r
JOIN Bi_DimCenter c
    ON r.centerid = c.centerid
GROUP BY c.centername
ORDER BY last_refresh DESC;

This query shows the latest refresh timestamp per center, helping you verify that all centers are receiving regular updates. The refresh log identifies a center by its centerid, so the query joins Bi_DimCenter to show the center's name. centerrefreshdatetime records the refresh time for the center; the log also has a refreshutcdatetim column -- spelled this way -- for the time in UTC.

Checking data freshness on any table

Most tables include ETL metadata columns that indicate when records were last processed:

Column Description
etlCreatedDate Timestamp when the record was first written by the data refresh process
EtlLastUpdatedOn Timestamp when the record was last updated by the data refresh process

Which of these columns a table has varies. Dimension tables usually carry EtlLastUpdatedOn and fact tables usually carry etlCreatedDate, so check the table's column reference (linked from Understand the star schema and data model) first. To check how recent the data in a table is, query the maximum value of the column it carries:

-- A fact table: when did it last receive new records?
SELECT MAX(etlCreatedDate) AS latest_load
FROM Bi_FactInvoiceItem;

-- A dimension table: when was one of its records last updated?
SELECT MAX(EtlLastUpdatedOn) AS latest_update
FROM Bi_DimCenter;

To identify records that were processed in the most recent refresh cycle, filter on etlCreatedDate:

SELECT *
FROM Bi_FactInvoiceItem
WHERE etlCreatedDate > '2025-12-01 00:00:00'
  AND void = 0;

Use these columns to build watermark-based processing in your downstream pipelines -- track the last etlCreatedDate you processed, then query for records newer than that timestamp.

Onboarding and initial data load

When your organization is first onboarded to Analytics — Redshift Data Share, the following process occurs:

  1. Zenoti provisions your access -- cluster credentials and endpoint details are provided to you
  2. Centers are enabled for analytics -- each center in your organization must be enabled for analytics data processing. Your Zenoti account team manages this during onboarding.
  3. Historical data is loaded -- the initial full load extracts and processes all historical data from your enabled centers. Depending on data volume, this may take longer than a regular refresh cycle.
  4. Incremental processing begins -- after the full load completes, the system switches to incremental mode, refreshing data approximately every 1--2 hours

Note: Centers must be enabled for analytics before their data appears in the data warehouse. If you add a new center to your organization after initial onboarding, contact your Zenoti account team to enable analytics for that center.

Automated scheduling

Pipeline execution is managed by an automated scheduling system. The scheduler coordinates processing runs across all stages of the pipeline, ensuring that each refresh cycle completes before the next one begins.

You do not need to trigger, schedule, or monitor refresh cycles. The system runs continuously and automatically. If a processing issue occurs, Zenoti's monitoring systems detect and address it.

Summary of key timing expectations

Aspect Expectation
Incremental refresh frequency Approximately every 1--2 hours
Initial onboarding Full historical load; timing depends on data volume
Consolidated table rebuild Every refresh cycle, using batch processing for large tables
Data lag Data reflects the state as of the last completed refresh cycle
Was this article helpful?
Your feedback helps us improve our documentation.