How does the VCS LDom Agent monitoring work?

Article: 100018572
Last Published: 2022-02-23
Ratings: 0 1
Product(s): InfoScale & Storage Foundation

Problem

How does the VCS LDom Agent monitoring work?

 

Solution

The VCS LDom (Solaris Logical domain) agent monitor script (if the MonitorCPU attribute is set) runs "ldm list-bindings -p <ldom>" command and parses the output to get the LDom status. This command returns output similar to this:
 
# ldm list-bindings -p LdomName
VERSION1.0
DOMAIN|name=LdomName|state=active|flags=normal|cons=5000|ncpu=3|mem=2147483648|util=100.0|uptime=621112
MAC|mac-addr=00:14:4f:f9:27:09
VCPU
|vid=0|pid=4|util=100|strand=100
|vid=1|pid=5|util=100|strand=100
|vid=2|pid=6|util=100|strand=100
MEMORY
|ra=0x8000000|pa=0x108000000|size=2147483648
VARIABLES
|auto-boot?=false
|boot-device=/virtual-devices@100/channel-devices@200/disk@0
VNET|name=vnet1|dev=network@0|service=primary-vsw0@primary|mac-addr=00:14:4f:fb:ab:b8
|peer=primary-vsw0@primary|mac-addr=00:14:4f:ff:ff:f1
|peer=vnet1@LdomName2|mac-addr=00:14:4f:ff:ff:f4
VDISK|name=vdisk1|vol=LdomName-bootvol@primary-vds0|dev=disk@0|server=primary
VCONS|group=Ldomgroup|service=primary-vcc0@primary|port=5000
 
As part of its function, the monitor script computes the CPU usage for each CPU and, if all of them are 100% busy, then it returns LDOM status as offline. If the LDom resource gets faulted too often, then increasing the ToleranceLimit to 3, for example, for LDom resources, will provide greater tolerance.

For further information on the LDom agent, please refer to the Cluster Server Bundled Agents guide, e.g. for 7.4.2:

https://sort.veritas.com/DocPortal/pdf/79620650-141543509-1

Was this content helpful?