Console system pages

Dashboard #

The dashboard presents the site’s current status, counts, and work queue. Technician tiles open jobs; the command bar searches device names. A failed summary read is not an all-clear: retry, then check System → Health if it still fails. An unreadable approvals queue is unknown; requests may still be waiting. Approve enrolled agents to enable patching, inventory, antivirus, and metrics on the host.

Maintenance windows suppress alarms until their configured end. Built-in detection continues without content packs. Check Detection packs for publishing and installation details; if publishing is unavailable, inspect Collector health.

The first-day checklist uses confirmed checks. A tick means the check succeeded; an unchecked result is unknown, not complete. Open each unfinished item to inspect it. Skip admission only when the site does not use network access control.

A disk failure can lose both appliance data and backups stored on that disk. Keep a current backup copy and an encrypted recovery kit away from the appliance. The protection warning clears only when both are confirmed.

Settings #

Read only and Permission required indicate that the signed-in role lacks the action’s capability. Ask an administrator to grant the required permission; details remain on the status tooltip.

Installation and repair #

The Windows appliance installer collects the appliance name, domain, and static IP settings before setup. Saved inputs and completed steps let a first installation resume after a required reboot. Opening settings on an installed appliance does not restart or reinstall it.

Repair uses the installed identity and preserves the existing domain controller. Correct the prerequisite named by the failed installer step, then resume with the same setup build. An administrator must schedule any restart required on an existing appliance; repair does not authorize an automatic reboot.

The setup receipt records the original site, connection profile, and password policy. Current password values come from the domain. A difference means that policy changed after the receipt. The receipt covers the password baseline; the checklist checks readiness, rather than attesting every installed setting still matches.

Branding #

Site branding overrides company branding, which overrides Skans defaults. Set the name, accent color, and logo address for the company or site. Skans vendor identity cannot be edited. The site option for the Skans credit overrides the company option. Cancel restores loaded values. Reset confirms and immediately saves Skans defaults. Save persists edited values, and Apply theme refreshes the console.

Retention #

Retention controls the age of logs, measurements, and activity records. Older data is removed during daily cleanup. Successful saves take effect at the next cleanup. The current values and defaults remain visible beside the form.

Pruning the activity log first copies entries older than the configured age to the off-box backup folder and verifies that copy. Without a folder or a confirmed copy, nothing is deleted. The scheduled cleanup job is SkansAuditRetention.

Retention days are a target. The disk budget takes priority: the smaller of the configured percentage, any GB cap, and available disk space sets the limit. Under pressure, older measurements are deleted before logs. Inventory and settings are not deleted by retention. The displayed estimate uses the current data rate and may differ from the requested days.

Automatic disk cleanup deletes search-service crash dumps and old backup snapshots while retaining the configured newest copies. It is enabled by default and records deletions. Preview cleanup lists planned deletions without changing data. The keep-count remains a lower bound for snapshot cleanup.

Detailed measurements remain for the configured retention period, then become hourly and daily summaries with their own retention periods. Summarization runs daily at 03:45 appliance time. Charts switch to summaries automatically. Summarize measurements runs this work on demand.

Outbound proxy #

An empty proxy host uses direct connections. Bypass entries also connect directly; a leading dot matches a host suffix. The proxy password is stored in the credential vault and is never displayed again. Save proxy persists the settings; Test checks the configured destinations.

Health #

Refresh measures appliance health again. Saved health samples expire after five minutes; their timestamps describe the last check, not a current live guarantee. Disk, memory, CPU, database, search, and service readings retain their individual status. Unknown or unavailable readings must not be treated as healthy.

Database recovery #

An unavailable database structure check means the check did not run. Its reason is available on the status; a reachable database at the next normal restart may allow verification. Do not restart a domain controller simply to clear this notice.

A database with no version label cannot be verified against the installed build. This can follow a restore. A previously untracked current database and an older restored database cannot safely be distinguished by assigning a new label. Rebuild creates a verifiable database and erases appliance data; use the confirmed recovery procedure.

For an older database, the page retains the actual and required version and unresolved changes. Startup adds missing objects where it can. Unresolved tables, columns, indexes, or foreign keys may prevent features from working. Check again reconciles and verifies the structure; a successful check clears the failure. If it remains, follow the rebuild procedure.

An unreadable settings file is refused rather than treated as empty. Its settings are not applied; enrollment, check-in, scan ranges, or maintenance windows may be affected. Reset settings renames the file rather than deleting it and starts with empty settings after confirmation. Re-enter the affected settings; preserve the original file for recovery.

Storage and other checks #

SQL Express limits each database’s data files to 10 GB; transaction logs are counted separately. A single size sample cannot establish growth. Storage metadata and certificate inventory records show their measurement time. A recorded certificate is inventory data, not a fresh connection to a device.

When search cannot be read, problem counts are unconfirmed. Open Services & roles, check the search service, and start it if stopped. If it is running, copy diagnostic details for support and refresh Health.

The encrypted secret store protects saved device logins and records access. Failed checks may mean saved logins are unavailable. Copy details for support before retrying. Network sensor counters are read directly from the appliance and remain available when search is down. They populate after device traffic or the daily self-test.

Certificates #

The certificate page shows appliance and device certificate state and expiry. Renew certificates approaching expiry. Device actions become available when their inventory rows have loaded.

Validity and renewal windows are displayed in days. Only Skans-owned device credentials explicitly enabled for rotation are changed at the configured rotation interval. Passwords shared with a recording system are never rotated.

Out of the box, device certificates last 365 days and renew themselves in the last 30 days; Skans-owned credentials enabled for rotation change every 90 days. These live in %ProgramData%\Skans\config\cert-policy.json (validityDays, renewWithinDays, rotateEveryDays).

Publish revocations makes the canceled-certificate list available for devices to check. Update sign-in certificate selects the current network sign-in certificate. Renew console certificate renews the console certificate and may restart the console; use its confirmation before proceeding.

The appliance uses one Enterprise Root certificate authority. Issuer and template values come from measured certificates. A dash means the value was unavailable. The revocation list retains its number, issue time, next update, and last publish outcome.

Expiring soon filters device certificates to the next 90 days. Clear the filter to see all devices. Issue certificate issues and sends a certificate to the selected device. Export CSV exports the current filtered device list.

Network #

The Network tabs cover name resolution, address assignment, admission, segmentation, configuration changes, device logs, and internal ranges. Network devices are listed under Devices.

Name resolution #

Test a name queries the appliance DNS server using either its LAN-client policy or its appliance policy. It does not contact the named device. Query failure details remain on the result tooltip.

DNS answers names for the appliance’s domain and sends other names to the configured forwarders. Add or correct records when a device name does not resolve. On an air-gapped site, upstream addresses can be recorded but are not used to forward queries until the site is declared connected.

For a reverse-zone PTR record, Name is the last part of the device IP address (for example 41); Data is its full host name. Forwarder addresses are comma separated. Old-record cleanup status and zone counts describe the resolver’s current configuration.

Address assignment #

DHCP scopes provide device IP addresses, DNS settings, and the domain. Run only one DHCP server per segment: activating an overlapping scope alongside a router or firewall DHCP server gives clients inconsistent settings. Disable the other server or leave the scope inactive.

Agents compare DHCP offers with the configured trusted server address. An empty override uses the appliance server. For a guest site using customer DHCP, set the customer’s address to keep rogue-server detection useful. The effective address remains visible. An unreadable scope list is unknown, rather than proof that no scopes exist. Approve server authorizes the appliance DHCP server in the domain.

Admission #

Switches and access points ask the network sign-in service whether a device may connect. Add each switch, configure its shared secret, point it at the site’s service, and enable certificate sign-in on its ports. The switch or access point enforces the decision; the server only answers requests.

Saving an existing switch updates its address. An empty shared secret keeps its existing secret. New secrets must meet the form’s minimum length. The allowed-switch list accepts comma-separated addresses or ranges; an empty list answers no switches. Its firewall rule follows the list after the next installer run. Segment values are 1–4094; zero leaves a segment unset.

The emergency open-ports action disables access control everywhere, including for unknown devices. Use its confirmation deliberately. Recent sign-ins cover 48 hours; Accept means allowed, Reject means denied.

Segmentation #

Appliance restrictions block selected appliance-to-device TCP administration ports. They do not isolate devices from each other or enforce gateway rules. Block remote access applies the appliance restriction for RDP, SMB, WinRM, SSH, Telnet, and RPC. Save rules exports a ruleset for a compatible switch or firewall. Apply that export on the gateway to enforce broader isolation. A listed device without an applied appliance rule is not yet restricted.

Configuration changes #

Configuration drift compares each switch, router, or firewall’s latest saved configuration with its last accepted copy. Accepting the latest copy changes that baseline; review unexpected differences first. Changes are recorded in the activity log. Configure backup schedules under Backup → Network device configs.

Device logs #

Devices send messages to the site’s Collector monitoring address: UDP 514 for syslog and UDP 162 for traps. Devices do not connect to the ControlPlane. Check receiving services when sources are missing. Search uses stored device messages; open Logs for full history and paging. A time-limited result is partial; narrow its time range in Logs. An unavailable message store leaves delivery and current results unconfirmed; retry and inspect the receiving services.

Internal ranges #

Private ranges are recognized automatically. Add other internal ranges, such as 100.64.0.0/10, with a reason. Skans treats those addresses as on site. Public ranges require confirmation. Enter one range per line or comma separate them; empty input clears the custom list.