Re: we have 3 Esxi node. tried adding 2nd and 3rd ...

amit_eric · ‎02-24-2019

we have 3 Esxi node. tried adding 2nd and 3rd node in cluster on 1. command was successful but when i checked cluster status all of them showing master locally which is not coorect i guess and Sub-Cluster Member Count is showing 1 in all 3 server.. if anybody have this kind of issue kindly help us. this is Esxi 6.5.

below is the output.

and steps follwed

ESXiHost1# esxcli vsan cluster new

ESXiHost1# esxcli vsan cluster get | grep 'Sub-Cluster UUID'

Sub-Cluster UUID: 52c284da-5978-3497-3f26-1d42a423b76

ESXiHost2# esxcli vsan cluster join -u <Sub-Cluster UUID noted on step 1>

ESXiHost3# esxcli vsan cluster join -u <Sub-Cluster UUID noted on step1 >

MGT1:~] esxcli vsan cluster get

Cluster Information

Enabled: true

Current Local Time: 2019-02-25T05:55:11Z

Local Node UUID: 5c6cca7c-e777-b6fd-9c86-48df373c177c

Local Node Type: NORMAL

Local Node State: MASTER

Local Node Health State: HEALTHY

Sub-Cluster Master UUID: 5c6cca7c-e777-b6fd-9c86-48df373c177c

Sub-Cluster Backup UUID:

Sub-Cluster UUID: 5269b313-4f65-9bb5-772b-17aa98b1454b

Sub-Cluster Membership Entry Revision: 0

Sub-Cluster Member Count: 1

Sub-Cluster Member UUIDs: 5c6cca7c-e777-b6fd-9c86-48df373c177c

Sub-Cluster Membership UUID: 74aa6f5c-0564-2e5e-eaaa-48df373c177c

Unicast Mode Enabled: true

Maintenance Mode State: OFF

Config Generation: 5c6cca7c-e777-b6fd-9c86-48df373c177c 1 2019-02-25T00:59:16.0

MGT2:~] esxcli vsan cluster get

Cluster Information

Enabled: true

Current Local Time: 2019-02-25T06:48:31Z

Local Node UUID: 5c6ce257-f09d-123b-06d7-48df373d3a34

Local Node Type: NORMAL

Local Node State: MASTER

Local Node Health State: HEALTHY

Sub-Cluster Master UUID: 5c6ce257-f09d-123b-06d7-48df373d3a34

Sub-Cluster Backup UUID:

Sub-Cluster UUID: 5269b313-4f65-9bb5-772b-17aa98b1454b

Sub-Cluster Membership Entry Revision: 0

Sub-Cluster Member Count: 1

Sub-Cluster Member UUIDs: 5c6ce257-f09d-123b-06d7-48df373d3a34

Sub-Cluster Membership UUID: 29ab6f5c-57df-01d1-c7d4-48df373d3a34

Unicast Mode Enabled: true

Maintenance Mode State: OFF

Config Generation: 5c6ce257-f09d-123b-06d7-48df373d3a34 4 2019-02-25T06:27:08.0

MGT3:~] esxcli vsan cluster get

Cluster Information

Enabled: true

Current Local Time: 2019-02-25T06:49:38Z

Local Node UUID: 5c6ce8ce-c8d7-f5ad-984d-48df373c1754

Local Node Type: NORMAL

Local Node State: MASTER

Local Node Health State: HEALTHY

Sub-Cluster Master UUID: 5c6ce8ce-c8d7-f5ad-984d-48df373c1754

Sub-Cluster Backup UUID:

Sub-Cluster UUID: 5269b313-4f65-9bb5-772b-17aa98b1454b

Sub-Cluster Membership Entry Revision: 0

Sub-Cluster Member Count: 1

Sub-Cluster Member UUIDs: 5c6ce8ce-c8d7-f5ad-984d-48df373c1754

Sub-Cluster Membership UUID: 9280735c-039d-80dc-50e2-48df373c1754

Unicast Mode Enabled: true

Maintenance Mode State: OFF

Config Generation: 5c6ce8ce-c8d7-f5ad-984d-48df373c1754 2 2019-02-25T06:15:13.0

TheBobkin · ‎02-25-2019

Hello amit_eric

Welcome to Communities.

All hosts are on same build version, on-disk format version and have a vSAN-enabled vmk configured on same subnet? Cluster Object has adequate vSAN license?

Any particular reason you are not using the web client wizard for cluster creation? This tells you what is misconfigured at many points where CLI won't.

Bob

amit_eric · ‎02-25-2019

Hi Bob,

just to add you, this was a re installation.

bob- >All hosts are on same build version, on-disk format version and have a vSAN-enabled vmk configured on same subnet? Cluster Object has adequate vSAN license?

Amit - Yes for all above points

bob - Any particular reason you are not using the web client wizard for cluster creation? This tells you what is misconfigured at many points where CLI won't.

Amit -> yes i agree,we have an instruction to do it through CLI only for installation and confiruation perspective,, one done we can configure rest of the things through GUI.

let me know your opinion.

Second is ,

last time i tried to clear partition and mapped the disk group with respective host and then tried to join the cluster , but due to node is not able to join the cluster(for which i have created this thread) we want to do it freshly.

we are thinking to wipe out everything and start fresh from installing Esxi and , can you tell me the step for clearing everything .

TheBobkin · ‎02-25-2019

Hello Amit,

"just to add you, this was a re installation."

There could be stale entries in CMMDS and/or disks still with partitions on them tying them back to the previous cluster - if there is no data on this cluster then validate there are no partitions on the vSAN disks (cache and capacity) before trying to join the cluster and remove existing vSAN partitions (Host > Configure > Storage Devices > Select disk > Other Actions > Erase partitions) make sure you are selecting the correct devices of course.

"yes i agree,we have an instruction to do it through CLI only for installation and confiruation perspective,, one done we can configure rest of the things through GUI."

I would not advise this, maybe back in the day (e.g. 5.5) there wouldn't be so much of a difference but for initial configuration vCenter is mandatory if you want everything to be clean - ideally the only time you should be doing this from the CLI should be for troubleshooting purposes when vCenter is not available. Scripted en-masse configurations (e.g. via PowerCLI) should go through vCenter too of course.

"we are thinking to wipe out everything and start fresh from installing Esxi and , can you tell me the step for clearing everything ."

You *shouldn't* have to wipe ESXi, just any existing vSAN partitions and configuration settings - e.g. place hosts in MM with 'No Action',delete the Disk-Group(s), ensure all partitions have been removed, move the hosts out of cluster Object into Datacenter Object, create new cluster Object, move hosts into it and then configure vSAN via the client.

Bob