Skip to content

Administration Audit and Diagnostics

bifrost0x edited this page Aug 14, 2026 · 1 revision

Administration, Audit, and Diagnostics

WebSSH's administration interface combines account lifecycle controls, runtime diagnostics, security auditing, host-key visibility, and native backup operations. Administrative access is necessary but is not a substitute for recent reauthentication on sensitive actions.

Administrative responsibilities

Administrators can:

  • create, enable, disable, and remove local accounts;
  • assign or remove administrator privileges;
  • inspect runtime and connection diagnostics;
  • export security audit records and update retention settings;
  • manage native backup and restore operations;
  • review host-key state and security-relevant configuration;
  • link or unlink supported external identities under the applicable safeguards.

Disabling or deleting a user revokes live activity, including Socket.IO, SSH, pooled connections, and transfers. Deletion first quarantines the user's data directory and rolls back that move if the database operation fails.

Recent authentication

Sensitive operations require a recent local-password confirmation. This limits damage from an unattended authenticated browser session. External identities do not bypass this safeguard; maintain a controlled local administrator credential and recovery process.

Security audit log

Security-relevant actions are written to DATA_DIR/logs/security_audit.log. Rotation uses AUDIT_LOG_MAX_BYTES and AUDIT_LOG_BACKUP_COUNT; defaults are 10 MiB and five backups.

Audit records are intended for investigation and accountability. They do not contain passwords, authentication tokens, private-key material, or file contents. Protect the log directory because event metadata can still reveal usernames, hosts, IP addresses, and administrative activity.

Audit export

The administrator export endpoint applies server-side filtering and scans at most 50,000 records per request. If a result is incomplete, WebSSH marks it as truncated in metadata and response headers. Narrow the time range or filters instead of assuming a truncated export is complete.

Retention updates are authenticated administrator operations and are themselves auditable. Rotation and retention protect local disk capacity; forward logs to an external system if the threat model requires tamper-resistant or longer-term retention.

Live diagnostics

Diagnostics expose bounded information about:

  • admitted Socket.IO connections;
  • active SSH and quick connections;
  • transfers and temporary storage;
  • background executor activity;
  • runtime shutdown or maintenance state;
  • configured capacity limits.

Use diagnostics to compare demand with GUNICORN_THREADS, socket admission, SSH quotas, transfer quotas, and background workers. Avoid treating a single counter as the entire capacity picture.

Host-key administration

Host-key trust is stored per user. First-use trust, changed-key rejection, and explicit revocation are covered in SSH Connections and Host Keys. Administrators should not globally bypass a user's trust decision to resolve a connection error.

Operational health

  • GET /health proves process liveness.
  • GET /ready checks whether the instance can accept work.

Readiness fails during maintenance, shutdown admission closure, database failure, or a failed data-directory probe. Details are in Health Checks and Troubleshooting.

Administration checklist

  1. Keep at least one tested local administrator and recovery path.
  2. Require HTTPS and a trusted reverse proxy for Internet exposure.
  3. Review disabled accounts, linked identities, and host-key changes regularly.
  4. Export or forward audit data according to the required retention period.
  5. Test native backups and restores on a separate instance.
  6. Monitor capacity and temporary disk usage before raising quotas.
  7. Apply upgrades only after reviewing configuration changes and release notes.

Clone this wiki locally