Display Absolute Number of Unprocessed Messages in Cluster Overview

What?

In previous versions of Graylog, the cluster overview at:

https://FQDN/system/cluster

provided a quick and convenient way to see the number of unprocessed messages for the cluster with a single click.

With the new cluster overview format, this information is no longer directly visible at the cluster level. The number of unprocessed messages can only be seen by opening and inspecting the individual node.

It would be very helpful to restore this information directly in the cluster overview by adding a dedicated column to the cluster table that displays the absolute number of unprocessed messages for each node.

For example:

Node | Unprocessed Messages
-- | --
graylog-node-01 | 12,458
graylog-node-02 | 0
graylog-node-03 | 847

The value should ideally represent the actual number of unprocessed messages rather than only displaying a percentage.

Why?

For Graylog administrators, the cluster overview is an important place to get a quick understanding of the current state of the cluster.

Having to open each individual node to determine whether there are unprocessed messages makes this process unnecessarily cumbersome and reduces transparency compared to the previous interface.

Additionally, displaying only a percentage is not an ideal replacement for the absolute number. For example, a very large journal could make a relatively small percentage represent a substantial number of messages.

The absolute number provides much more useful operational information when assessing the actual backlog.

A dedicated "Unprocessed Messages" column in the cluster overview would make it possible to identify nodes with a backlog immediately, without having to inspect each node individually.

Would it be possible to add the actual number of unprocessed messages natively to the cluster overview table?

Your Environment

• Graylog Version: 7.1.8
• OpenSearch Version: 2.19.4
• MongoDB Version: 8.2
• Operating System: Ubuntu 24
• Browser version: Brave latest

Start at the /system/cluster cluster overview and inspect how the cluster table gets its node data. Trace where the existing unprocessed-message percentage or node details are exposed, then add the requested absolute-count column. Done means each node shows its unprocessed message count directly in the overview, including zero, with coverage for the table behavior.

Written by the indexing model from the issue text.