Troubleshooting

Cygna Forest Recovery is built to say what went wrong and what to do about it. Every refusal, failed job, and preflight check carries a code and a sentence that names the cause, and most of them name the fix. Start with what the console tells you, and go to log files only when it does not say enough.

General approach

  1. Read the message in full. A backup job's What went wrong, a preflight check's remediation, and an alert's advice are written to be acted on. The code beside each one is stable and is what this chapter is organized by.

  2. Check that every component is connected. Most failures follow from an agent that is not connected or a storage agent that cannot be reached. The Agents screen shows the connection state and last heartbeat of every agent.

  3. Look at the transcript. Backups and restores keep transcripts, which the console shows: on the Backup Jobs detail sheet, and under What the machine reported on a restore run's page.

  4. Read the service log of the component involved, around the time of the problem. See Logs and Diagnostics.

  5. Check the audit trail for what changed: a setting, a policy, a role. See Auditing Operator Activity.

  6. Before you contact support, create a support package. See Collecting a support package below.

Where to look

Problem Start with
A backup failed or did not run The job on the Backup Jobs screen, and Backup Job Errors
A restore or recovery plan cannot be approved The Preflight findings card on a recovery plan's page, or the Preflight card on a restore run's page, and Preflight Check Reference
A restore stopped or failed The outcome and What the machine reported on the restore run's page, then the restore transcript on the domain controller
An agent does not appear, or shows as not connected The agent's service log, and Agent Enrollment and Connectivity
A push installation failed The deployment's per-target result, and Agent Enrollment and Connectivity
An installer refused to install Installer Messages
An alert is raised The alert's advice on the Alerts screen, and Alerts Reference
Full system recovery is unavailable for a domain controller The Recovery image column on the OS Media screen
The Manager is gone or will not start Recovering the Manager

Using the command line

The cdr command-line tool shows the same information as the console, and adds detail that is useful when diagnosing. cdr jobs list --failed --detail adds the staging path, the agent log path, and the full error to each failed job. cdr agent list --offline and --silent-for <minutes> find agents that have gone quiet. cdr config show --agents --behind lists machines that are not running the current configuration. --json on any command gives output you can save or send to support.

cdr also uses exit codes you can act on: 2 means the Manager could not be reached, 9 that its CA is not trusted, 10 that you are not signed in, and 3 that the Manager refused the request, which includes an operator lacking the capability. See Command-Line Reference.

Collecting a support package

When you contact Cygna Labs Corp. support, send a support package. It collects the Manager's log, a copy of the Manager database with every credential removed, host and configuration facts, and the logs, job transcripts, and state files of every connected agent, into one zip file. You need the config.export capability (System Admin).

  • In the console, use the Support package card on the License screen. See Support package.
  • From a command prompt, run cdr support package. See Support Commands.
  • If the Manager service does not start, run CygnaDR.Manager.exe support package on the Manager as an administrator. It collects the Manager's part only.

The package is customer confidential and not encrypted. For what it contains and leaves out, how long the Manager keeps it, and what else to send, see Logs and Diagnostics.

In this chapter