Configuration Guide

8.2.1 Alerting Configuration Guide

replication-manager can deliver alerts through multiple channels when cluster state changes. Alerts fire when a monitored error or warning code first appears in the cluster state and again when it clears.

8.2.2 Alert Triggering

monitoring-alert-trigger
Description Comma-separated list of error/warning codes that trigger an alert notification. Only codes in this list cause notifications to be sent — all other state changes are logged but do not fire the alert channels.
Type String
Default "ERR00027,ERR00042,ERR00087,ERR00002,WARN0023,WARN0100,WARN0115,WARN0116,WARN0139,WARN0140,WARN0141"
Example "ERR00027,ERR00042,WARN0100"
monitoring-ignore-errors
Description Comma-separated list of error or warning codes to suppress entirely. Matching codes are never logged and never trigger alerts. Use this to silence known false positives.
Type String
Default ""
Example "WARN0067,WARN0066"

8.2.3 Alert Blackout Window

Alerts can be suppressed on a schedule — useful for planned maintenance windows or known noisy periods (e.g. nightly batch jobs).

scheduler-alert-disable
Description Enable the scheduled alert blackout. When true, replication-manager silences all alert channels during the window defined by scheduler-alert-disable-cron and scheduler-alert-disable-time. Can also be toggled at runtime from the GUI.
Type Boolean
Default false
scheduler-alert-disable-cron
Description Cron expression (6 fields: second minute hour day month weekday) that defines when the alert blackout window starts.
Type String
Default "0 0 0 * * 0-4" (midnight, Monday–Friday)
Example "0 30 22 * * *" (22:30 every day)
scheduler-alert-disable-time
Description Duration of the alert blackout window in seconds.
Type Integer
Default 3600 (1 hour)

8.2.4 External Script

An external script is called on every state change for every monitored server, regardless of monitoring-alert-trigger. It receives the affected server information as arguments.

alert-script
Description Full path to a script to execute when a server state changes. The script is called with three positional arguments: server URL, previous state, current state.
Type String
Default "" (disabled)
Example "/opt/scripts/repman-alert.sh" 
#!/bin/bash
# Arguments: $1=server_url  $2=previous_state  $3=current_state
echo "$(date) $1 changed from $2 to $3" >> /var/log/repman-alerts.log

8.2.5 Email

Email alerts require a reachable SMTP server. The mail alert fires for all state transitions.

mail-from
Description Sender address for alert emails. Must be a valid user on the configured SMTP server.
Type String
Default "mrm@localhost"
mail-to
Description Comma-separated list of recipient addresses. Leave empty to disable email alerting.
Type String
Default ""
Example "[email protected],[email protected]"
mail-smtp-addr
Description SMTP server address in host:port format.
Type String
Default "localhost:25"
Example "smtp.example.com:587"
mail-smtp-user
Description SMTP authentication username. Leave empty for unauthenticated relay.
Type String
Default ""
mail-smtp-password
Description SMTP authentication password.
Type String
Default ""
mail-smtp-tls-skip-verify
Description Connect to SMTP with TLS but skip certificate verification.
Type Boolean
Default false
mail-max-pool
Description Maximum number of persistent SMTP connections to keep open. 0 disables connection pooling (a new connection is opened for each alert).
Type Integer
Default 0
mail-timeout
Description Timeout in seconds when waiting for an SMTP connection from the pool. 0 means no timeout.
Type Integer
Default 5

8.2.6 Slack

Slack alerts use an incoming webhook URL. Slack alerting is global to the replication-manager instance — all clusters share the same webhook.

alert-slack-url
Description Slack incoming webhook URL. Leave empty to disable Slack alerting.
Type String
Default ""
Example See Slack → App Management → Incoming Webhooks for your workspace URL
alert-slack-channel
Description Slack channel to post alerts to.
Type String
Default "#support"
alert-slack-user
Description Display name for the Slack bot posting the alert.
Type String
Default ""

8.2.7 Pushover

Pushover delivers mobile push notifications to iOS and Android via the Pushover service.

alert-pushover-app-token
Description Pushover application token (created at pushover.net). Leave empty to disable Pushover alerting.
Type String
Default ""
alert-pushover-user-token
Description Pushover user or group key — identifies who receives the notification.
Type String
Default ""

8.2.8 Microsoft Teams

Teams alerts are sent via an incoming webhook. Alerting can be filtered to specific severity levels.

alert-teams-url
Description Microsoft Teams incoming webhook URL. Leave empty to disable Teams alerting.
Type String
Default ""
alert-teams-proxy-url
Description HTTP proxy URL to use when calling the Teams webhook (for environments where outbound HTTP requires a proxy).
Type String
Default ""
alert-teams-state
Description Comma-separated list of severity prefixes to forward to Teams. Only state changes whose code starts with one of these prefixes are sent. Leave empty to send all triggered alerts.
Type String
Default ""
Example "ERR" (errors only) · "ERR,WARN" (errors and warnings) · "ERR,WARN,INFO" (all)

8.2.9 Syslog

log-syslog
Description Mirror all log output to the local UDP syslog port in addition to the normal log file. Allows log aggregation systems (Graylog, Splunk, ELK) to receive replication-manager events via syslog.
Type Boolean
Default false

8.2.10 Cloud18 Integration

When the instance is registered with the Signal18 SSO, alert notifications can also be forwarded to the Signal18 Mattermost instance. These settings are global (not per-cluster).

cloud18-alert
Description Enable forwarding of alerts to the Cloud18 Slack/Mattermost channel.
Type Boolean
Default true
cloud18-alert-slack-channel
Description Mattermost channel name to post Cloud18 alerts to.
Type String
Default "signal18_alert"
cloud18-alert-slack-url
Description Webhook URL for the Cloud18 Mattermost instance.
Type String
Default "https://meet.signal18.io/hooks/…"
cloud18-alert-slack-user
Description Display name for the Cloud18 alert bot.
Type String
Default "repman"

8.2.11 Checking Alert Status via API

The following endpoints allow external monitoring systems to poll cluster health:

GET /api/status
# {"alive": "running"} or {"alive": "starting"}

GET /api/clusters/{clusterName}/status
# {"alive": "running"} or {"alive": "errors"}

GET /api/clusters/{clusterName}/topology/alerts
# Returns the current active alert list

Via CLI:

replication-manager-cli status
# running

replication-manager-cli status --cluster=my-cluster
# running  or  errors

replication-manager-cli status --cluster=my-cluster --with-errors
# JSON object listing all active errors and warnings