Using fence_scsi + cmirror together in a RHEL 5 cluster can result in a fenced node getting stuck on boot and blocking write I/O to mirrors across the cluster

Solution Unverified - Updated -

Issue

  • When a node that was fenced with fence_scsi reboots, it gets stuck on boot-up with SCSI reservation conflicts and path failures. All write I/O issued by other nodes in the cluster is stuck until that node is fenced again
  • After a node is fenced, on boot the cluster services (including scsi_reserve) appear to start up OK, but reservation failures commence immediately on the newly booted node

Environment

  • Red Hat Enterprise Linux (RHEL) 5 with the High Availability Add On
  • SCSI Persistent Reservation Fencing (fence_scsi)
  • Cluster-Mirrored LVM2 volumes (cmirror)
  • Physical volumes in mirror backed by device-mapper-multipath maps

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.

Current Customers and Partners

Log in for full access

Log In
Close

Welcome! Check out the Getting Started with Red Hat page for quick tours and guides for common tasks.