Monitors

Path: Left sidebar > Infrastructure > Monitors

Monitor Overview

The Cluster Monitors page shows the health and status of all Ceph Monitor (MON) daemons. Monitors maintain the cluster map - the authoritative record of cluster topology, OSD states, and PG mappings.

What You See

The Monitors page is split into three panels:

  • Status (left)

  • In Quorum (right top)

  • Not In Quorum (right bottom)

Purpose:

  • To verify monitor quorum before performing storage operations.

  • To confirm the control-plane state is stable.

When to Use:

  • Before creating or changing pools, file systems, or block/object resources.

  • During any alert about monitor health or quorum.

Steps:

  1. Open Infrastructure > Monitors.

  2. Check In Quorum panel. All deployed monitors should appear here.

  3. Check Not In Quorum panel. Healthy state shows All monitors are in quorum..

  4. Check Status panel quorum field and confirm expected monitor count.

Expected Outcome:

  • You confirm whether quorum is healthy and the cluster can safely accept writes.

Infrastructure Monitors dashboard

What This Screenshot Shows: Monitors Dashboard (UI Reference; Values Depend On Your Environment).

Status Panel - Field Reference

Field

What It Shows

Cluster ID

Unique cluster FSID shared by all daemons

monmap modified

Timestamp of last monitor map change

monmap epoch

Monitor map version number (increments on map changes)

quorum con

Quorum-supported connection feature set

quorum mon

Reserved field (often -)

required con

Required connection feature set (often -)

required mon

Required monitor feature set (often -)

quorum

Monitor names currently in quorum with count

In Quorum Panel - Column Reference

Column

What It Shows

Name

Monitor daemon identifier (for example mon.hostname)

Rank

Numeric rank assigned during election. The leader holds Rank 0.

Public Address

IP address and port this monitor is listening on

Open Sessions

Number of currently open client sessions on this monitor

Tip

Rank 0 is the current leader monitor coordinating map updates and elections. Frequent leadership changes can indicate monitor network instability.

Tip

Open Sessions indicates active monitor client connections. Temporary variation is normal; sustained imbalance can indicate client routing or host-level issues.

Monitor Roles Reference

Role

Meaning

Leader

The primary monitor that coordinates map updates and elections

Peon

A follower monitor that participates in quorum but defers to the leader

Probing

A monitor attempting to establish quorum - not yet fully connected

Quorum

The cluster requires a majority of monitors to form quorum before it can accept client I/O.

  • For a 3-monitor deployment, at least 2 must be in quorum.

  • For a 5-monitor deployment, at least 3 must be in quorum.

Warning

If quorum is lost, the cluster will stop accepting writes. Restore monitor connectivity immediately.

Expanded Row Details

Field

Description

FSID

The unique cluster identifier. All daemons in the cluster share the same FSID.

Election Epoch

Increments on every leader election. Frequent elections indicate network or latency issues between monitors.

Map Epoch

The current version of the monitor map.

Features

Ceph feature flags supported by this monitor daemon version.

Warning

Rapidly increasing election-related counters indicate unstable monitor connectivity. Validate monitor network latency and time synchronization.

CLI Diagnostic Context

The CLI commands in the Common Issues table are optional diagnostics for administrators.

Before running them:

  • SSH to the monitor host.

  • Use an account with sudo/administrative privileges.

  • Keep UI workflow as primary path for normal operations.

If this fails

Symptom

Likely Cause

Action

Monitor not in quorum

Network partition or daemon crash

Check ceph mon stat and journalctl -u ceph-mon@<id>

Leader election loop

Clock skew between hosts

Sync NTP on all monitor hosts

1 of 3 monitors missing

Host unreachable

Restore host or replace monitor

Cluster is not accepting writes

Quorum lost (below majority)

Restore connectivity until quorum majority is re-established

Note

Operational note: Deploy monitors on an odd number of dedicated hosts (3 or 5) with synchronized clocks and low-latency network connections.

Note

If any issue persists, raise a support ticket via Monitoring > Alerts or Karios Support.