Skip to main content
Migration Notice
We're migrating documentation from the old portal into this one. Some things may look a little different or out of place in the meantime — we know, and we're working to get it right. If something's unclear or doesn't look right, let us know.
Version: 2.15.1

Release Notes for Data Security for Dell

This page lists the changes, fixes, and known issues for Data Security for Dell version 2.15.1.

What's Fixed in Version 2.15.1​

  • License Management Cards Show the Correct Capacity: Fixed a capacity display issue on the License Management cards, where some licenses, such as Disaster Recovery, showed a capacity of 0 / 0 TB.

  • Disk Usage Is Now Shown for ECS Devices in Inventory: On Inventory → Manage Devices, ECS devices previously showed an empty disk usage slot. Disk usage is now reported for ECS devices. Consumed space is calculated as total minus free, so space reserved by ECS counts as consumed and the figure can read higher than the "% Full" shown in the ECS console. Devices that have not yet reported inventory show an "awaiting inventory" state instead of an empty value.

  • Custom Queries Can Be Scheduled in the New GUI: Scheduling a custom query in the New GUI failed with a message that the schedule could not be attached. Daily, Weekly, and Monthly schedules now attach correctly and run at the hour, minute, and day you select.

  • Unusable Event Types Removed from the Data Auditing Event Type Picker: FILE_SET_SEC, DIR_SET_SEC, and ADMIN_RESYNC were listed as Event Types in both the New and Old GUIs, but PowerScale never emits them, so selecting one always returned an empty report. They are no longer offered, which reduces the list from 23 to 20 event types. Security changes are reported as FILE_SET_ACL and DIR_SET_ACL.

  • ECA Firewall Now Blocks Unauthorized Network Traffic by Default: The ECA host firewall's final drop rule only applied to an interface named eth0, so on appliances whose network interface uses a different name (such as ens160), traffic from hosts outside the allowed list was not blocked and internal cluster services could be reached from arbitrary network hosts. The firewall now denies all traffic that is not explicitly allowed, regardless of the interface name. SSH and HTTPS access are unchanged.

  • Restoring a Pre-2.15 Backup Onto a 2.15 Appliance No Longer Stops sca.service: Restoring a backup taken on a release earlier than 2.15 onto a 2.15 appliance could leave sca.service unable to start, with a keystore password verification error, because the restored keystore password was not carried over to the new 2.15 password location. The restore now reconciles the password, and the service starts normally after the restore.

  • Zookeeper Now Starts With the Configured Log Level: After a restart or cluster up, Zookeeper could start in debug mode instead of using the configured log level, which could contribute to memory issues on the ECA. Zookeeper now starts with the intended log level.

  • /opt/data No Longer Fills Up With Zookeeper and Turbo Audit Logs: Fixed an issue where the Zookeeper and turboaudit log directories under /opt/data/superna/logs kept growing until /opt/data reached 100% utilization, which left ECA nodes inactive.

  • Zookeeper Backup and Restore Problem Fixed: Fixed a problem with the Zookeeper backup and restore process.

Known Issues​

The following known issues have been identified in version 2.15.1.

Security Guard​

  • Security Guard for ECS Fails When the ECS Object User Name Contains an Underscore: Security Guard creates a bucket named igls-security-guard-bucket-<username>, and bucket names cannot contain an underscore. When the ECS object user name includes one, such as the default ECS user object_admin1, the Security Guard job fails to create the bucket.

    Workaround: Use an ECS object user name that has no underscore or other special characters. Delete the Security Guard bucket that was already created, edit the user name used for Security Guard, and run Security Guard again.

  • No Option to Configure an ECS Object Access User for Security Guard: Neither the Old (legacy) GUI nor the New GUI provides a field to configure an ECS Object Access User on the Security Guard page. In the Old GUI, Security Guard cannot be configured until at least one ECS Object Access User is set.

  • Security Guard for ECS Run Steps Show Out-of-Order Timestamps: In the New GUI, the timestamps of all Security Guard steps are preset to the start time of the run. While a run is in progress, a completed step can show a later time than the steps that follow it.

  • Security Guard for ECS Fails When Many Previous Security Guard Files Exist: When the ECS has a large number of Security Guard files from earlier runs, each run removes only part of them and then fails. It can take several runs before the remaining files are few enough for a run to succeed.

    Workaround: Run Security Guard again until it completes.

  • The Restore Object User Access Step Takes 15 Minutes: The Restore object user access step of an ECS Security Guard run always takes 15 minutes.

Health Check​

  • Removing a License Does Not Remove Robo Audit from the Health Check Overview Page: After removing the Easy Auditor license, Robo Audit is removed from the settings page but still appears on the Health Check Overview page. This applies to the PowerScale Security Guard license.

  • Time Skew Alarm Is Raised for a Late ECA Heartbeat: Eyeglass raises the alarm Time skew between ECA and the Appliance is too great (EAU0004) when it handles an ECA node heartbeat more than 5 seconds after the node sent it, even if the clocks of the ECA node and Eyeglass agree. This can happen while Eyeglass is busy, for example during an upgrade restart or heavy use of the UI or API. You can receive bursts of Warning alarms and alarm emails for one or more ECA nodes, followed by no alarms. Checking the clocks shows no difference.

Threat Detections​

  • Closing a Threat Detection in the Old GUI Does Not Restore a Locked-Out User: In the New GUI, the close wizard tells you to restore the user's access before you close the threat. In the Old (legacy) GUI, closing the event does not restore the access of a locked-out user.

    Workaround: In the Old GUI, restore the user's access before you close the event.

Active Auditor​

  • Active Auditing Is Not Available to the Auditor Role in the New GUI: Users with the Auditor role can access Active Auditing in the Old (legacy) GUI. In the New GUI, Active Auditing is part of Threat Detections and is not available to the Auditor role.

Previous Release - 2.15.0​

What's Fixed in Version 2.15.0​

  • Snapshot History Now Correctly Reflects Reused Snapshots for Events Raised While Pending: A Ransomware Defender event that had a snapshot job run while still in Pending Events, and was later promoted to Active Events, now correctly shows the reused snapshot against the event, and the event's history correctly records the snapshot activity. Previously, the event's snapshot list and history showed no record of the snapshot at all, even though the correct snapshot had been reused rather than duplicated.

  • Reused Snapshots Are Now Verified Before Reuse: Eyeglass could previously still treat a snapshot as reusable after it had been deleted directly on OneFS or through the event history's delete-snapshot action, as long as it was still within the reuse window — so a new event wouldn't get a real snapshot, and attempts to extend the (deleted) snapshot's expiry would fail with an error. Eyeglass now verifies a candidate snapshot still exists on the cluster before reusing it, discarding the stale record and creating a new snapshot if it was deleted outside of Eyeglass.

Known Issues​

The following known issues have been identified in version 2.15.0.

General​

  • Signing Into One UI Signs You Out of the Other: Signing in to the Old (legacy) GUI while already signed in to the New GUI — or vice versa — signs you out of the other interface after a few seconds. Signing in to the New GUI first does not cause this and then signing in to the Old GUI.

Licensing​

  • Licensing a Managed Device for Data Security May Show a "Cannot Connect" Error: Licensing in the new Eyeglass UI is product-family based, and Data Security covers both Easy Auditor and Ransomware Defender. If a managed device has only one of the two licensed, applying a Data Security license to it fails for the unlicensed application and shows a "Cannot Connect" error.

    Workaround: Ignore the error if the other application is already licensed correctly. Otherwise, assign the missing license in License Management and re-run inventory.

Active Auditor​

  • Active Auditor DLP Trigger Shows Incorrect Threat Category: A Data Loss Prevention (DLP) trigger event's threat category currently displays as "Suspicious Activity" in the event slideout, instead of the correct "Data Loss Prevention."

  • Active Auditor Allows Creating a CRTA Policy with Empty Audit Criteria: In the new UI, a Custom Real-Time Audit (CRTA) policy can be created and saved with empty audit criteria, with no validation warning shown. The previous UI correctly blocked this and displayed an "audit criteria is empty" warning; the new UI is missing this validation, allowing an invalid or incomplete policy to be created.

  • Active Auditor Events With Multiple Triggers Show Duplicate "Suspicious Activity" Labels: When an Active Auditor event is triggered by more than one rule at once (for example, both Mass Delete and a Custom Trigger), the event slideout and Investigation page show the threat category as repeated, generic "Suspicious Activity" entries instead of naming each distinct trigger. The Signal Strength panel already identifies the actual triggers involved.

Threat Hunting​

  • Threat Hunting Instance Cannot Be Deleted from Manage Services: Attempting to delete a Threat Hunting instance from Manage Services fails with an error stating the service could not be found at the specified IP and port, and the instance is not removed.

Threat Detections​

  • Threat Detections Slideout Window Does Not Auto-Refresh: The Threat Detections event slideout window does not consistently refresh in real time. Fields such as Items Affected and lockout time in the details section may not update automatically after an action is taken; the slideout must be closed and reopened to see the current data.

Integrations​

  • Webhook Payload URL Field Does Not Link to the Correct Event: The URL field in the Data Security webhook payload sent to integrated SIEM/SOAR solutions does not link to the correct event. A fix is planned for version 2.16.0.

Investigate & Recover​

  • Items & Recovery CSV Download Tooltip and Status Inaccuracies: On the Investigate & Recover page's Items & Recovery tab, the download tooltip label is incorrect, and only a CSV download option is available (no Download XML option). Additionally, the downloaded CSV file always shows "Recoverable" as the status for every row, regardless of the item's actual recoverable, unrecoverable, or recovered state.

Detection Controls​

  • Some Multi-Word Custom Suspicious Extensions Cannot Be Removed in the New GUI: A custom suspicious extension entry made up of two words cannot always be removed from the Suspicious Extensions list in the New GUI.

  • Learned Thresholds Entry Shows a Path Under User ID After Marking an Event as False Positive: When a threat detection event is marked as a false positive, the resulting Learned Thresholds entry incorrectly includes a path under the User ID field. This does not happen in the Old GUI.

  • Warning Expiry Field Can Appear Blank on First Load (Old GUI): In the Old (legacy) GUI, the WARNING severity Threshold's Expiry (minutes) field can appear blank the first time the Threshold panel is opened after a page reload, even though a value is actually saved. Reopening the panel a second time shows the correct value. If changes are saved while the field displays blank, the blank value is accepted and can overwrite the real expiry setting.

  • A Note Can Be Lost When Converting an Ignore List Entry to Monitor Only (Old GUI): If a path already has a Monitor Only entry, converting an Ignore List entry for that same path to Monitor Only silently drops the incoming note rather than merging it — the pre-existing Monitor Only note is kept instead.

Data Auditing​

  • WireTap Events Show the Same Value in the Path and File Columns: In the New GUI, a WireTap event row displays identical data in both the Path and File columns. The Old (legacy) GUI shows only a Path column. A fix is planned for a future release.

  • Robo Audit Report Discoverability and Timestamp Inconsistencies: When a Robo Audit job runs on schedule or manually, the job appears on the Health Check Runs page, but there's no clear indication that the generated user/path audit reports can be found under Data Auditing's Queries & Reports "All Reports" tab. On the All Reports tab, it's difficult to determine which user/path audit report corresponds to which Robo Audit job, and the report's listed start time reflects the Robo Audit job's start time rather than the individual user/path query's start time.

    Workaround: Compare the Robo Audit job's timestamp on the Runs page with the timestamps on the All Reports page. Note that the Runs page displays time in 24-hour format while All Reports uses 12-hour format, and that All Reports shows the Robo Audit job's start time rather than the individual user/path query's start time.

See Also​

Need Help?​

  • Contact Support: Reach out to Superna Support for configuration assistance or troubleshooting help.