[LU-8457] Pacemaker script to monitor LNet

Blocker — Wed, 26 Apr 2017 11:07:30 +0000

A new script to be used in Pacemaker to monitor LNet compatible with ZFS and LDISKFS based Lustre server installations.

This RA is able to monitor a single LNet device using the Pacemaker's clone technology.

pcs resource create [Resource Name] ocf:pacemaker:healthLNET \
dampen=[seconds 5s] \
multiplier=[number 1000] \
lctl=[ true | false] \ 
device=[device name ib0] \ 
host_list=[ list of NIDs, space separated, if lctl is true otherwise list of IPs] \
--clone

where:

dampen The time to wait (dampening) further changes occur
multiplier The number by which to multiply the number of connected ping nodes by
attempts Number of ping attempts, per host, before declaring it dead
timeout How long, in seconds, to wait before declaring a ping lost
lctl Option to enable lctl ping instead of the normal ping. The default is true
device Device used for the LNET network. We assume the same device accross the cluster

This script should be located in /usr/lib/ocf/resource.d/heartbeat/ of both the Lustre servers with permission 755.

Default values:

dampen 5s
multiplier 1
attempts 3
timeout 5s
lctl true

Default timeout:

start timeout 60s
stop timeout 20s
monitor timeout 60s interval 10s

Compatible and tested:

pacemaker 1.1.13
corosync 2.3.4
pcs 0.9.143
RHEL/CentOS 7.2

Example of procedure to configure:

pcs resource create healthLNET ocf:pacemaker:healthLNET dampen=5s multiplier=1000 lctl=true device=eth1 host_list="10.10.130.1@tcp1 10.10.130.2@tcp1" --clone 

targets=`crm_mon -1|grep 'OST'| awk '{print $1}'` 

for i in $targets; do pcs constraint location $i rule score=-INFINITY pingd lt 1 or not_defined pingd; done

Whamcloud Community JIRA

[LU-8457] Pacemaker script to monitor LNet