18Monitoring and logs
Last updated
SYNERGY Management Gateway records what it does in several places, each meant for a different reader and a different question. This chapter covers the tools for watching the platform itself: the live application log, the log files on the server, forwarding the log to an external system or to Sentry, exporting traces and metrics to an OpenTelemetry collector, and the health endpoints that monitoring tools and load balancers can poll.
These tools describe the platform. They are not the record of who changed what in Active Directory. That record is the operations log, described in Chapter 9, Operations log, approvals and waiting operations.
Overview
The table below summarises the monitoring tools, who can use them and what they are for.
| Tool | Where | Who can use it | Purpose |
|---|---|---|---|
| Live Logs | Monitoring › Live Logs | System Admin role only | Watch the application log as it is written; read the log files on the server |
| Log Stream | Monitoring › Log Stream | System Admin role only | Forward the application log to a SIEM, collector, automation or Sentry |
| Telemetry | Monitoring › Telemetry | telemetry:read / telemetry:manage, Observability licence module | Export traces and metrics over OTLP |
| Health endpoints | /health/live, /health/ready | Anyone (no sign-in) | Liveness and readiness probes |
The Monitoring group in the navigation menu (shown as Observability when the menu is expanded) also contains Operations and My Waiting Operations; those are covered in Chapter 9.
Important: Live Logs and Log Stream are available only while your active role is the System Admin role. No grantable permission unlocks them. The application log can contain server internals such as LDAP filters, exception text and stack traces, so it is treated as system-level information. If you hold System Admin among several roles, switch to it first (see Chapter 3, Signing in and your account).
Application log basics
Every component of the platform writes to one application log. Each entry has:
- Time – when the entry was written.
- Level – how serious it is (see the table below).
- Source – the source context, which is the name of the component that wrote the entry, for example
Synergy.ManagementGateway.ActiveDirectoryfor directory operations orSynergy.ManagementGateway.Infrastructure.Notificationsfor email notifications. - Message – the text of the entry.
- Exception – for errors, the technical details of the failure, including a stack trace.
The levels, from least to most serious, are:
| Level | Meaning |
|---|---|
| Verbose | Everything the platform writes. Normally not recorded at all. |
| Debug | Diagnostic detail, very high volume. Normally not recorded. |
| Information | Normal activity, including every request the platform serves. Hundreds an hour on a busy instance. |
| Warning | Something went wrong but the platform recovered. A handful a day on a healthy instance. |
| Error | A failure. |
| Fatal | The platform could not continue. |
By default the server records Information and above. Verbose and Debug entries are not produced unless the server's minimum level is lowered in its configuration file (the Serilog:MinimumLevel section of the API's appsettings.Production.json). This affects every log tool in this chapter: you cannot see or forward entries the server never writes.
In this chapter
Last updated
Thank you for your feedback.