Incremental (Delta) Syncs: Overview and How They Work

What an Incremental Sync Is

Lumos keeps a full copy of your identity and account data in sync with your source systems, such as your HRIS and directory. Historically, that data was refreshed through a complete sync that runs on a set schedule (for example, once per day or every few hours). A complete sync reads the entire population of accounts each time it runs. This keeps everything accurate but takes longer and cannot run frequently for large data sets.

An incremental sync is a lighter, faster sync that runs in between your complete syncs. Instead of reading every account, it pulls only the records that have changed since the last sync. Because it fetches only deltas, it completes much faster and can run far more often. The goal is to shorten the time between a change happening in your source system and that change appearing in Lumos.

In short: complete syncs guarantee the full picture on a slower cadence, and incremental syncs keep that picture fresh in between by capturing just what changed.

What Incremental Syncs Capture

Incremental syncs focus on account and identity data. They pick up:

• New accounts (for example, a new hire created in your HRIS or directory)

• Status changes (for example, an account moving to a terminated or inactive state)

• Attribute changes (updates to identity attributes that Lumos tracks)

• Account deletion events

The following data is not included in incremental syncs and continues to be refreshed by your regular complete syncs:

• Entitlement data (group memberships, roles, permissions)

• Last activity / usage data

This is why complete syncs remain part of the picture. Incremental syncs keep account and identity attributes current throughout the day, while the complete sync refreshes the full data set, including the categories above.

How Often They Run

Incremental syncs run on a configurable interval (for example, every two hours, or more frequently if the changed-data pulls are quick). The cadence is set based on how your specific integration behaves, including how long each sync takes to complete. Faster, lighter pulls can support shorter intervals; heavier ones need more spacing.

During the current phase, incremental sync schedules are set up and adjusted by Lumos rather than self-service in the product. To change the cadence, contact Lumos.

How Incremental and Complete Syncs Work Together

This is the most important behavior to understand, because it explains how the two sync types interact in practice.

For a given integration, an incremental sync and a complete sync will not run at the same time. If a complete sync is already running for that integration, the next scheduled incremental sync is skipped until the complete sync finishes. The two take turns rather than overlapping.

This has a practical consequence worth planning around. If your complete sync runs frequently and takes a long time to finish, it can occupy windows where an incremental sync would otherwise have run. That reduces the number of incremental syncs that actually execute in a day. For example, if a complete sync runs every four hours and takes close to three hours to complete, incremental syncs scheduled for every two hours will only have a clear window roughly every other cycle.

The way to get the most value from incremental syncs is to give them room to run. That typically means spacing out the complete sync (running it less often, since incremental syncs now keep data fresh in between) and tightening the incremental interval. When incremental pulls complete quickly, they can be scheduled at short intervals to capture changes such as new hires and status updates throughout the day. The complete sync then runs on a slower background cadence to refresh everything else.

Which Integrations Support Incremental Syncs

Incremental sync support is enabled on a per-integration basis, because each source system has to support returning only its changed records. It is not a global setting that applies to every connected application at once.

Supported integration types currently include Active Directory, Workday, Okta, JDBC, and flatfile-based connectors. Support continues to expand to additional systems over time. If there is a specific source system you would like to enable incremental syncs for, contact Lumos to confirm current availability or the expected timeline.

Summary

• Incremental syncs capture only changed account and identity data, so they run quickly and frequently in between your complete syncs.

• They cover new accounts, status changes, attribute changes, and deletions. Entitlements and activity data still come from the complete sync.

• They run on a configurable interval set by Lumos, not on an exact clock time.

• For any given integration, incremental and complete syncs do not run simultaneously, so spacing out the complete sync and tightening the incremental interval produces the freshest data.

• Support is enabled integration by integration and currently includes Active Directory, Workday, Okta, JDBC, and flatfile connectors.