In a Pacemaker cluster, as in a standalone system, operators must complete maintenance tasks such as software upgrades and configuration changes. Here’s what you need to keep Pacemaker’s built-in monitoring features from creating unwanted side effects.

Maintenance mode

This is quite possibly Pacemaker’s single most useful feature …

If you’ve heard of Ceph, you’ve surely heard of radosgw, a RESTful gateway interface to the RADOS object store. You’ve probably also heard that it provides a front-end interface that is compatible with Amazon’s S3 API.

The question remains, if you have an S3 client that …

For users of VMware virtualization, it’s becoming increasingly common to deploy Pacemaker clusters within the virtual infrastructure. Doing this requires that you set up fencing via ESX Server or, more commonly, vCenter. Here’s how to do that.

The cluster-glue package contains node Pacemaker’s fencing (STONITH) plugins, one …

Ever wonder what’s the difference between order <name> inf: <first-resource> <second-resource> and a score of something other than inf? We’ll explain.

If you specify an order constraint score of INFINITY (inf or the keyword mandatory in crm shell syntax), then the order constraint is considered mandatory. If you …

It’s not uncommon in Pacemaker clusters to run specific cron jobs only on a node that currently runs a particular resource. The ocf:heartbeat:symlink resource agent can be exceptionally helpful in this situation. Here’s how to use it.

Suppose you’ve got a cron job for Postfix …