VMware ESXi 6.7 U2
I just upgraded my ESXI hosts 4 to 6.7 U2
All hosts have and SSD Flash drive and 3 HDD drives of the same size
I turned on VSAN which then showed all disks and I claimed them all it then went on to process the VSAM configuration.
One Host was not communicating and the process timed out. The Host was non responsive from the console the VMS were still accessible.
I had to power off the host and powered back on all came back. the VMS are now all up and running.
On all four Hosts I have this message
Host cannot communicate with one or more other nodes in the vSAN enabled cluster
On the Cluster I have these messages.
vSAN health alarms are suppressed
vSAN datastore vsanDatastore in cluster TGCSNET-Vcenter2-Cluster in datacenter Datacenter-TGCSNET-2 does not have capacity
When I look at the datastores the VSAN-Datastore has no disks
Any ideas or thoughts on how to get this running?
Thank you
Tom
host 16 is part of same vsan cluster.? If yes, please remove from it and re-add host back to the cluster. Previous my answer was based on your first question.
please mark helpful or correct if issue resolve.
Tom, you really need to spend some time in reading the documentation before starting with this product. This isn't something you just "wing." That said, you have a networking issue. Have you created vmkernel ports explicitly for vSAN and tagged them for this service? This is most likely a basic issue that could be understood when studying this product before attempting to implement it.
sorry I did not get back so soon the site keeps telling me it is in maintenance mode.
I have a network setup just for VSAN
I ran this on host 1
[root@TGCSESXI-15:~] esxcli vsan cluster get
Cluster Information
Enabled: true
Current Local Time: 2020-03-18T22:34:20Z
Local Node UUID: 5d38d9dd-590c-374e-20f8-001018f421ec
Local Node Type: NORMAL
Local Node State: MASTER
Local Node Health State: HEALTHY
Sub-Cluster Master UUID: 5d38d9dd-590c-374e-20f8-001018f421ec
Sub-Cluster Backup UUID: 5d3bf7da-6354-159a-dbb8-000af7a054c8
Sub-Cluster UUID: 52be451c-96dc-7a62-3e6f-9adf883c5415
Sub-Cluster Membership Entry Revision: 4
Sub-Cluster Member Count: 3
Sub-Cluster Member UUIDs: 5d38d9dd-590c-374e-20f8-001018f421ec, 5d3bf7da-6354-159a-dbb8-000af7a054c8, 5d3e54c3-dad7-0f94-1d59-001018f511a0
Sub-Cluster Member HostNames: TGCSESXI-15.our.network.tgcsnet.com, TGCSESXI-17.our.network.tgcsnet.com, TGCSESXI-18.our.network.tgcsnet.com
Sub-Cluster Membership UUID: 4b74725e-7982-fae5-9e54-001018f421ec
Unicast Mode Enabled: true
Maintenance Mode State: OFF
Config Generation: 4d3e48f2-e56c-4f5f-bb29-c3b16c0be4dd 3 2020-03-18T20:39:00.353
Then on Host 2 which is the host that froze
[root@TGCSESXI-16:~] esxcli vsan cluster get
vSAN Clustering is not enabled on this host
How can I enable vsan on one of my hosts? The other three look ok
Once the Node 2 gets configured I believe the vsan will config everything
Or will I need to run a command to start the process again?
Thank you
Tom
Workaround:
Restart the VPXA management agent by running this command
/etc/init.d/vpxa restart
Restart hostd and vsanmgmtd service in ESXi.
/etc/init.d/hostd restart
/etc/init.d/vsanmgmtd restart
If service restart does not resolve the issue, place the host in maintenance mode and reboot.
Please mark correct or helpful if resolve the issue.
host 16 is part of same vsan cluster.? If yes, please remove from it and re-add host back to the cluster. Previous my answer was based on your first question.
please mark helpful or correct if issue resolve.
Nawals,
Moving it out of the cluster and then back to the cluster worked it added the vsan configuration.
[root@TGCSESXI-16:~] esxcli vsan cluster get
Cluster Information
Enabled: true
Current Local Time: 2020-03-19T12:17:01Z
Local Node UUID: 5d3b6dc2-19e8-f2da-5ba4-000af77acb40
Local Node Type: NORMAL
Local Node State: AGENT
Local Node Health State: HEALTHY
Sub-Cluster Master UUID: 5d38d9dd-590c-374e-20f8-001018f421ec
Sub-Cluster Backup UUID: 5d3bf7da-6354-159a-dbb8-000af7a054c8
Sub-Cluster UUID: 52be451c-96dc-7a62-3e6f-9adf883c5415
Sub-Cluster Membership Entry Revision: 5
Sub-Cluster Member Count: 4
Sub-Cluster Member UUIDs: 5d38d9dd-590c-374e-20f8-001018f421ec, 5d3bf7da-6354-159a-dbb8-000af7a054c8, 5d3e54c3-dad7-0f94-1d59-001018f511a0, 5d3b6dc2-19e8-f2da-5ba4-000af77acb40
Sub-Cluster Member HostNames: TGCSESXI-15.our.network.tgcsnet.com, TGCSESXI-17.our.network.tgcsnet.com, TGCSESXI-18.our.network.tgcsnet.com, TGCSESXI-16.our.network.tgcsnet.com
Sub-Cluster Membership UUID: 4b74725e-7982-fae5-9e54-001018f421ec
Unicast Mode Enabled: true
Maintenance Mode State: OFF
Config Generation: 4d3e48f2-e56c-4f5f-bb29-c3b16c0be4dd 5 2020-03-19T12:15:45.884
Now I just need to get the VSANDatastore with capacity.
Any way to trigger that to process again?
Thanks Tom
I figured it out.
Was able to claim all the SSD as cache and all HDD as capacity
The VSAN Datastore now has capacity great.
When I went to create a folder I got this message
This operation requires 2 or more usable fault domains.
I have one fault domain with all my hosts assigned.
Any ideas.
Thanks Tom
sorry I figured this out also.
I do not require a Fault domain since they only work with hosts in the same rack
I deleted the Fault domain and now able to write to the VSANDatastore.
Thanks Tom
Follow below to fix to get vSandatastore with capacity.
Workaround:
Restart hostd and vsanmgmtd service in ESXi.
/etc/init.d/hostd restart
/etc/init.d/vsanmgmtd restart
If service restart does not resolve the issue, place the host in maintenance mode and reboot.
Related VMware KB:
VSAN cluster would display incorrect capacity information with 0 Bytes used and 0 Bytes free (58928)
https://kb.vmware.com/s/article/58928
Nawals
I posted another question.
After several attempts to get VSAN to work I gave up on this My ESXI host keep freezing up after access the vsan datastore from my veeam replication jobs.
I decided to remove vsan all together but now my local drives are read only I can not create a new datastore from them
I posted another question in the VSAN group if you have any ideas on how a can get my drives formatted so I can make them local datastores.
Thank you
Tom