ESXi 6.5 not responding

[juansg]

Hi all,

I have an ESXi 6.5 host Lenovo Flex System appears as disconnected in VMware vCenter Server

Connecting to the ESXi host directly using the vSphere Client fails

Can not execute esxcli commands

Connection Only by ssh

I request help to recover the administration of the node without affecting the vm on.

Regards

[daphnissov]

You can try restarting vpxa and hostd, but if that fails grab logs and you may have to kill the host.

[matze007]

You can try to migrate the VMs to another Cluster / Host with a fling tool without turning the VMs offline.

Cross vCenter Workload Migration Utility

After it, you can restart the "error" Host.

[daphnissov]

That is most likely not going to work. If the host is disconnected then so are its VMs. The Fling will not be able to migrate them in that case.

[matze007]

Yeah you are right, so just you said, try to restart the service vpxa and hostd.

You can find more information about the agent that communicates with vCenter Server in the /var/log/vpxa.log

[juansg]

I have not restarted my esxi 6.5 host, since I found new errors, someone has had something similar:

I have a datastore with the name of my host and it is inaccessible.

-Anyone will have an answer, without affecting the vm's running

Thank you

[daphnissov]

You probably have a storage related failure and your VMs may already be down. Did you check?

[juansg]

My vm's are stored in an independent datastore, apparently the error is in local storage.

I have a question, if I restart my host esxi 6.5 lift or no longer have access to anything. And if I reinstall it, I recover my vm's?

Regards

[daphnissov]

if I restart my host esxi 6.5 lift or no longer have access to anything. And if I reinstall it, I recover my vm's?

I don't understand this question.

[juansg]

In reference to the errors of the previous image, with a reboot I can recover the administration of the esxi at the vcenter level and the web interface?, Or how can I recover the node, which I suggest, to that error is due.

The VMs are running without problem, mark errors of vSphere HA agent is not reachable, Can not synchronize host, Connection timed out and from the console of esxí, it sends me the error: nmp_pathdeterminefailure: 3335: scsi cmd reserve failed on path vmhba0

Regards

[daphnissov]

Yes, if you reboot the host, you should regain management capabilities. As I believe we said earlier, the first thing to try is to restart hostd and/or vpxa agents. If you aren't using things like LACP, vSAN, or the like, you can just do a services.sh restart from an SSH session. The storage-related errors you're seeing may or may not be the reason why this has hung. If you need a definitive answer, you should gather logs and open a SR with VMware.

[Vikas_VM]

1.  If you are able to take the console form IMM, then try to restart the Management Services (if SSH is not working).

2. If VM's ae on shared datastore then you can shutdown the VM's and reregister to other host.