Skip to main content
AI Data Engine

Configure data classification thresholds and access scanning in AI Data Engine

Contributors netapp-dbagwell
PDFs

Cold and stale thresholds help you find storage that you can optimize using AI Data Engine (AIDE). Permission scanning and Active Directory integration help you find and understand access risks. Set them to match your organization's data classification, retention, and access policies.

AI Data Engine general settings

Before you begin

You must have the AI Data Engine administrator role.

What is cold and stale data?

Cold data

Files that haven't been accessed within the cold data threshold. Cold data is a key metric your organization can use to identify storage that could be optimized according to your retention and business policies.

Stale data

Files that haven't been modified within the stale data threshold. Stale data helps your organization identify files that haven't changed recently, which you can monitor or manage according to your policies.

These thresholds affect dashboard summaries, the files shown when you apply cold or stale filters in Investigation, and the files that policies target. Choose thresholds that reflect how long data can go unused or unchanged before your organization evaluates it. A shorter threshold classifies more data as cold or stale, and a longer threshold narrows the results to the data that has been inactive the longest.

Note The default threshold values are product defaults provided as a starting point. They don't represent legal, compliance, or industry-standard retention periods. Set thresholds that reflect your organization's specific policies and requirements.

Changes to the cold and stale thresholds take effect after the next scan cycle completes.

Set the cold data threshold

Decide how long files can go without being accessed before your organization evaluates them for archiving, deletion, or monitoring. You can also limit the classification by file size, and keep specific file types or paths out of it.

Steps

  1. From the AI Data Engine UI, select Configuration > General settings.

  2. Expand Cold data definition, and enable cold data classification if it is disabled.

  3. Under Age based on, select the timestamp that determines age: Last accessed to identify files that haven't been used recently, or Last modified to identify files that haven't changed recently.

  4. Set the Cold age threshold. The default is one year.

  5. Set the Minimum file size (MiB) and Max file size cap (MiB) to focus on the files that matter to you.

  6. Optional: To keep specific files out of this classification, enter Exclude file types and Exclude paths. Enter files and full paths, and separate multiple entries with commas.

  7. Select Save.

Set the stale data threshold

Decide how long files can go without changing before your organization reviews them under its data lifecycle policies. You can also limit the classification by file size, and keep specific file types or paths out of it.

Steps

  1. From the AI Data Engine UI, select Configuration > General settings.

  2. Expand Stale data definition, and enable stale data classification if it is disabled.

  3. Under Age based on, select the timestamp that determines age: Last modified or Last accessed.

  4. Set the Stale age threshold. The default is three years.

  5. Set the Minimum file size (MiB) and Max file size cap (MiB) to focus on the files that matter to you.

  6. Optional: To keep specific files out of this classification, enter Exclude file types and Exclude paths. Enter files and full paths, and separate multiple entries with commas.

  7. Select Save.

Find access risks with permission scanning

Turn on permission scanning when you need to find files and folders that are open to more people than your policies allow. To review the results, refer to Investigate directories.

Steps

  1. From the AI Data Engine UI, select Configuration > General settings.

  2. Expand Permission scanning, and turn on the toggle.

    To stop scanning permissions, turn the toggle off.

Show user and group names with Active Directory

Connect AI Data Engine to Active Directory when you need to see user names and permissions, rather than numeric IDs, as you review your data. If your environment uses NFS, you can also map numeric user IDs (UIDs) and group IDs (GIDs) to directory users and groups.

For the prerequisites and steps, refer to Configure Active Directory identity mapping for NFS.