Have a question? Want to report an issue? Contact JFrog support

Skip to end of metadata
Go to start of metadata

Overview

You can view the status of, and manage your HA cluster nodes in the Admin module under Configuration | High Availability. 

This screen displays a table with details on all the Artifactory nodes in your cluster as displayed below:

Page Contents

The table columns are as follows:

ID
Unique descriptive name of the server.
Start Time
The time that the server was started .
URL
The context URL that should be used to communicate with this server within the cluster.
Membership Port
The port that should be used to communicate with this server within the cluster.
State

The current state of the server as follows:

  • Offline - The node has started in an invalid state (For example, same HA license, or the same node id has been set on two different nodes). In this case you should the specific server logs for details.
  • Starting - The node is starting up.
  • Running - The node is up and running. This is the normal state for a node.
  • Stopping - The node is in the process of shutting down.
  • Stopped - The node is shut down.
  • Converting  - The node is converting database tables and configuration files.
Role

The role of the server as follows:

  • Primary - The primary node.
  • Member - A regular member node.
  • Standalone - The node is not configured into your HA cluster. It is running as a separate installation of Artifactory (Pro or Open Source).
Last Heartbeat
The last time this server signaled that it is up and running. By default, each node signals every 5 seconds.
Version
The Artifactory version running on this cluster node.
Revision
The Artifactory revision number running on this cluster node.
Release
The Artifactory release running on this cluster node.

Monitoring for Unresponsive Nodes

You can use the Last Heartbeat field to identify unresponsive nodes. If a node abruptly stops working (e.g. system crash on the server), then it may not be able to correctly update its State value, and will continue to appear as Running. 

However, since the server Heartbeat does not get updated for a long interval of time, it is displayed in red with a warning sign as displayed below.

In this case you should check that the corresponding server is up and running and fully connected to your HA cluster and the database.

HA unresponsive node


Removing an Unused Node

If you remove a node from your cluster, it will still appear in your database and will therefore be displayed in the list of cluster nodes.

To avoid confusion you should remove it from your list of cluster nodes (and detach it from the database) so that it doesn't interfere with normal cluster behavior.

To do so, hover over the corresponding server from the list and click "Delete".

Removing an unresponsive node

When to remove a node

The "Remove" button is only available once a node has not signaled a Heartbeat for a "long" time.

We recommend that you only remove a node from your list if it has indeed been removed from your system.

In case of an error in one of the nodes, or if a node is shut down for maintenance, the Heartbeat will not be updated and Artifactory will alert you to this as described in the previous section.

In this case there is no need to remove the node from your list. Once you fix the error and restart the server the Heartbeat will be updated and the warning will be dismissed.

 

 

  • No labels