Bug ID 4406523

This bug was previously reported incorrectly as Bug ID 4480277.

Problem Summary: This bug reported two problems for VERITAS Volume Manager, the first of which has been corrected by a code fix. In the second problem reported, during volume recovery a volume manager (VM) slave node that attempts to join the cluster might cause the VM master node to crash. "VM slave node" means any node that leaves the cluster as the result of a panic.

Workaround: To prevent the VM slave node from causing the VM master node to crash while the master node is synchronizing the shared volumes, perform the following two steps.

1. Update the /etc/system file of each cluster nodes to include the following line, then reboot each node to make the change take effect.

   set halt_on_panic=1

2. On all cluster nodes, change the eeprom auto-boot? flag to false.

   eeprom auto-boot?=false

   Some shell systems use slightly different syntax. See the man page for the eeprom(1M) command for the exact syntax to use for your shell system.