VMworld 2018: Hardware Health Monitoring in vSphere Stack

VMworld 2018 saw a lot of exciting announcements made by both VMware and Dell Technologies. One of the key breakout sessions at VMworld 2018 that got significant customer interest was the topic titled "Changes to Hardware Health Monitoring in vSphere Stack" by Michael MacFaden and Chaitanya Gangwar from VMware. DellEMC works closely with VMware to simplify the vSphere hardware management experience with the results from our collaboration featured in this session. These important enhancements in vSphere systems management stack have been introduced in ESX6.7 and ESX6.5U2. 

ESXi systems management framework enables customers to manage and monitor the Host hardware infrastructure. Dell EMC PowerEdge servers are powered by iDRAC to enable and support management software such as OpenManage Server Administrator (OMSA), iSM, ESXi and vCenter to monitor and manage host infrastructure. The key highlights from this breakout session were the enhancements in the Host and vCenter user interface, hardware events and alarms.

Host Client interface and vCenter User interface

• The Host client user interface now presents a more refined and purposeful hardware details compared to a relatively static representation of the host infrastructure.
• Michael highlighted how the ESX6.7 and ESXi6.5U2 web client interface organizes the sensor data based on severity reported by iDRAC rather than presenting a static listing base on sensor type.
• Additionally, the ESXi host client interface now integrates the relevant System Event Log (SEL) information for each sensor. This change is very important since it enables a system admin to quickly get a comprehensive view of server health based on platform sensor status. Server storage details are now separated out from the IPMI sensor data and presented in a new "Storage" tab under Monitor->Hardware. 
• vSphere Client also has the capability to show the Server hardware status for each of the hosts it manages. It’s usually found under <Host> -> Monitor->"Hardware Health".  
• A lot of changes have gone in to make sure that relevant sensor details are displayed providing a more accurate state of the hardware. A side effect of this could be a reduced number of sensors shown in vCenter 6.7 compared to earlier releases, but this is primarily because vSphere client now doesn’t list components such as software packages and sensors that don’t represent health status.

ESXi6.7 and ESX6.5u2 introduces a new stateless sensor specific event type, which complements the existing stateful group event type. The existing group events keeps the Virtual Center view of the host hardware state in synchronization with ESXi hardware health. Group event categorizes events/alarms based on common sensor type such as processor, memory, fan etc. Having an additional stateless event type that considers individual IPMI sensor state change vs a group change helps eliminate false positive scenarios while reporting hardware health of an ESXi host.

vSphere APIs and CLIs for reporting Hardware Health
Finally, this breakout session highlighted on how MOB (Managed Object Browser) can be used to explore hardware management APIs. This might be of interest to customers who would like to use vSphere APIs in their custom data center management software. For a more native experience, ESXCLI has also been enhanced to support additional IPMI commands.