@HockeyFan04, Regarding disks being failed/offline I meant vSAN disks (e.g. Cache-tier/Capacity-tier devices), again this will be shown in the Health UI (or in Disk Management UI - Cluster > Configure > Disk Management) but you can also determine which disk/Disk-Groups/nodes are the problem by looking at a few Inaccessible Objects and noting where the Absent/Degraded components reside (e.g multiple Inaccessible Object's Absent/Degraded components having particular disk/Disk-Group/nodes in common).
By cluster partitioned I was asking are some nodes not communicating with other nodes in the cluster e.g. they are in their own network partition.
Disk Balance in warning state is not of consequence with regard to data availability.
With regard to MM - While they not show vSphere MM being enabled in the UI (e.g. the host icon is different) but is important to validate that they are also out of vSAN MM (for which there is a Health check if they are not in sync)
https://kb.vmware.com/s/article/51464
What build version of ESXi and vCenter is installed?
Did any changes to the cluster precede this state? (e.g. disks or nodes physically removed/replaced, nodes put in MM or rebooted, nodes updated/upgraded, network changes etc.)
On any node in the cluster (provided the cluster is fully-formed - validate this with node count from 'esxcli vsan cluster get') run the following command to identify how many Object are in which unhealthy state e.g. state: 12 and state: 13 are two of the expected Inaccessible states (remove all leading '#' of course):
# cmmds-tool find -f python | grep CONFIG_STATUS -B 4 -A 6 | grep 'uuid\|content' | grep -o 'state\\\":\ [0-9]*' | sort | uniq -c
Then run the following 2 commands (or whichever is applicable e.g. if all Inaccessible are state: 12 then only the 1st one, if a mix of 12 and 13 then both etc.) and either share or PM me the output:
# for i in `cmmds-tool find -f python|grep FIG_STAT -C6|grep 'uuid\|content'|grep 'state\\\": 12' -B1|grep uuid|cut -d "\"" -f4`;do echo;echo "DOM_UUID: $i";echo;cmmds-tool find -f json -t DOM_OBJECT -u $i|grep content |sed 's/,/\n/g'|grep -E "componentUuid|componentState|faultDomain|diskUuid|type"|grep -v StateTS|sed 's/\}//g;s/{//g;s/"//g;s/content: //g;s/,//g;'|sed 's/attributes: //g'|grep -vE "Configuration|RAID"|sed 's/Witness/Witness /g' |sed 'N;N;N;N;s/\n/ /g';echo;python -c "print('*' * 124)";done
# for i in `cmmds-tool find -f python|grep FIG_STAT -C6|grep 'uuid\|content'|grep 'state\\\": 13' -B1|grep uuid|cut -d "\"" -f4`;do echo;echo "DOM_UUID: $i";echo;cmmds-tool find -f json -t DOM_OBJECT -u $i|grep content |sed 's/,/\n/g'|grep -E "componentUuid|componentState|faultDomain|diskUuid|type"|grep -v StateTS|sed 's/\}//g;s/{//g;s/"//g;s/content: //g;s/,//g;'|sed 's/attributes: //g'|grep -vE "Configuration|RAID"|sed 's/Witness/Witness /g' |sed 'N;N;N;N;s/\n/ /g';echo;python -c "print('*' * 124)";done
I am not aware of how good this forum platform (Khoros) is for not introducing line-breaks so do validate each command is one-line only.
