All ESXi related objects (vm, datastore, etc) are now their own sensors. Using options UI, you can configure how their state is presented. See documentation for details
VM sensors now include guest IP address (as attribute)
License sensor now include host (as attribute) to which they are attached
VM sensor cpu_use_% value changed to display 2 decimal places
YAML is no longer supported, please use Integrations UI
This release is a breaking change, and hopefully the last one. In the future releases Iāll try to have the component handle breaking changes on its own.
If configured via Integrations UI
Update the component
Restart HASS - at this point, the component will fail to load. Thatās ok
Once HASS is back, go to Configuration > Integrations > ESXi Stats integrtion and remove it
I donāt see a way to set the card size within the configuration of this card, maybe some other custom card can be used here. I personally have all my VM stuff in its own tab with panel: true. In a horizontal stack I include 6 cards which scale to a size that is comfortable for me.
Icon configuration - for this to properly work with this card, you need to have the state of the sensor show information you want the icon to reflect.
For example, if you want the icon color based on VM running state, you need to have the Virtual Machine State Attribute set to āstateā (this is the default). Then you would change the decluttering template to the following. I added comments to show what I changed.
This will set the icon color to green if the VM is running, yellow if the VM is suspended, and grey if the VM is off.
If you want to display icon color based on VM status (green/yellow/red/grey) you would need to change the Virtual Machine State Attribute to āstatusā and then modify the automation section of the card to match possible states, like this:
i change the attribute to āstateā and add the code
Where did you change this? In the decluttering template or in the Integrations UI options? It needs to be set to state in Integrations UI options, like so:
Do you happen to have cpu_use_% as a VM State attribute set in options? Can you check your Integration Options (Configuration > Integrations > ESXi Stats) and share what you have set there?
I canāt say for certain since those fields were blank when I went back to review the settings. I suspect the blank field is what triggered all of the repetitive warnings. That said, configuring the missing setting and restarting HA resolved the issue.
It appears that one of my previously reported errors is still present in the home-assistant.log file. There appears to be some sort of value mismatch as the sensor is expecting a number but is getting a string, in this case green.
This occurs on both of my Windows Server 2016 VMs but none of the other OS VMs, including Windows 7.
[homeassistant.helpers.condition] Value cannot be processed as a number: <state
sensor.esxi_vm_dc01=green; name=dc01, status=green, state=running,
uptime_hours=3858.9, cpu_count=1, cpu_use_pct=0.89, memory_allocated_mb=4096,
memory_used_mb=4133, used_space_gb=43.71, tools_status=toolsOk,
guest_os=Microsoft Windows Server 2016 (64-bit), guest_ip=<redacted>, snapshots=0,
unit_of_measurement=Status, friendly_name=ESXi vm dc01 @ 2019-12-
22T17:05:48.460155-08:00> (Offending entity: green)
I need to take a deeper dive into this - at first glance it looks like HA doesnāt like the value of āgreenā for the sensor output and is expecting a number? Maybeā¦ but if it did, Iād expect that same behavior for all the sensors not just windows VMs. Also not sure if the error is introduced in 0.5.1 ESXi stats release or if some change in latest HA release. Iāll try to replicate it and go from there.