No connectivity between one machine and some (not all) other addresses

[cawst]

I've got an ESXi 6.5 host on a Cisco C220 M4 server, and it's connected to a Catalyst 2960-X stack.  The host is on VLAN 11, and the guests are on VLAN 22 (two different vSwitches, and different physical ports on the Catalysts).  This has all been working fine.  However, yesterday I tried to move my vCenter server onto this host, and suddenly the networking went crazy.  The guests can still ping everything normally, but the host cannot ping a variety of address on both VLAN 11 and 22.  But others on each VLAN it can ping just fine.  I cannot find a pattern to the ones it can or cannot reach (note: this is in comparison to what another previously identically set-up host can reach, so these should be responding without issue).  I've done everything I can figure out to do on the ESXi side - double checked all the networking setup, restarted everything, etc...  And on the Catalysts it all looks normal and like the other hosts as well.

I'm assuming the problem is on the host itself, but are there any tests I can do on the stack to absolutely confirm it's not being blocked in there, that it's definitely on the host?  I'm completely stumped, and so far haven't had any ideas come in from posting on the VMWare forum, so just trying to see what other avenues I have for tracking this weird issue down.

[Jaderson Pessoa]

Dear cawst,

Try this on vmware vswitch, check on their configuration on security tab if "promiscous mode" is enabled, if no, enable it please. And check your connections again.

To more information how to do it: https://kb.vmware.com/s/article/1004099

[cawst]

It was disabled.  I enabled it, but the issues persists.

[Jaderson Pessoa]

does your vswitch is configured as access mode or trunk? and your switch how is configured?

[cawst]

It's a port group on the Catalyst, and so it's set up as VLAN 0 on the vSwitch, same as the working hosts.  None of this has changed since it worked.

[Jaderson Pessoa]

Mayber your por groups has a wrong configuration your having a network loop..

could you provide a configuration from your por group? and if possible

show interfaces trunk

[cawst]

I'll have to dig that out when I'm back in the office.  However, how/why do you think it might be looping when pinging one address on a VLAN but not another?  And the same for the other VLAN?  Also, as mentioned, the config is identical to a host that's not having these issues, so are there other factors that could be contributing?  Is there some sort of test I could do to cut any variables out?

[Jaderson Pessoa]

Well, we need working with your information. But you can check port speed, duplex, erros with damaged cable, that you can have a packet loss, giant, hunt... There are many things, but you can start with : show interface gi1/0/1 (for exemple) and check if having a problem with, giant, crc, interface reset. Because if there, maybe its a damaged cable.

Without information we just can work with possibility.