This week I had the pleasure of deploying 2 more Nutanix blocks on behalf of one our partners, who is now starting to highly recommend Nutanix for their customer deployments of critical systems.
The installation was pretty vanilla, 3 NX-1065-G5 nodes at the Primary site and matching at the DR site. For the VMware components, we went with the vCenter 6.5 appliance (I love the stability and speed of the 6.5 appliance by the way), and for the ESXi hosts we went with 6.5 (build 4887370).
The install went great, super fast and easy as always is the case with Nutanix deployments, and off we were rolling for customer deployment.
After running the command ncc health_checks run_all post install (running ncc version 3.0.4-b0379d15 for this), I noticed that the results were calling out 3 hosts for having disabled services.
Detailed information for esx_check_services:Node 10.xx.xx.xx: WARN: Services disabled on ESXi host: -- sfcbd-watchdogNode 10.xx.xx.xx: WARN: Services disabled on ESXi host: -- sfcbd-watchdogNode 10.xx.xx.xx: WARN: Services disabled on ESXi host: -- sfcbd-watchdog
After doing some research on why the sfcbd-watchdog wasn’t starting – and trying to start it manually, I came across this KBase from VMware, which detailed that this was expected behavior starting in ESXi 6.5.
Wondering if the NCC code just wasn’t updated for this specific change from VMware, I checked the Nutanix Knowledge Base, and came across this link which details that services identified by the ncc health_checks hypervisor_checks esx_check_services command should be enabled.
Ok, so that makes sense… ESXi 6.5 has been out long enough to assume that the ncc scripts have been updated to accomodate the 6.5 changes. So time to get the service re-enabled, and check ncc again.
To enable the service on a ESXi 6.5 host, use the command esxcli system wbem set –enable true (be sure to use double hypens!). Per VMware, if a 3rd party CIM provider is installed, sfcbd and openwsman should start automatically. Just to be safe I also ran /etc/init.d/sfcbd-watchdog start, followed by /etc/init.d/sfcbd-watchdog status to make sure my services started.
So let’s see what we get now after running the ncc checks again, using the command ncc health_checks hypervisor_checks esx_check_services to simplify my results.
Results look much better, no more warnings about disabled services on the ESXi hosts.
Running : health_checks hypervisor_checks esx_check_services[==================================================] 100%/health_checks/hypervisor_checks/esx_check_services [ PASS ] -------------------------------------------------------------------------------------------------------------------------------------------------------++---------------+| State | Count |+---------------+| Pass | 1 || Total | 1 |+---------------+Plugin output written to /home/nutanix/data/logs/ncc-output-latest.log
Good to know that VMware purposefully has disabled this service, and it’s easy to put that in a checklist for future deployments. I do wish though that since Foundations is taking care of the ESXi install and customization, they would add those 2 cli commands to the routine to make those services start, if they truly are needed.
Hope this helps if you run into this same issue!