What’s Core Dump ?

The network Core Dump functionality in ESXi helps in capturing diagnostic data through the network in the event of a purple diagnostic screen.

Configuring Core Dump

First, you need to start Core Dump from your vCenter. You could connect to your vCenter VCSA in SSH or use the web client with administrator@vsphere.local account to start Core dump service.

Start Core Dump from SSH :

# service-control --start vmware-netdumper

You could show with this command if your service is running.

# service-control --status vmware-netdumper

Start Core Dump from Web Client :

Click on start button to start this service.
For remember, you could find all services here (Administration > System Configuration > Services)

Configure Core Dump on each ESXi Host

You can configure Core Dump for each ESXi hosts in two different ways.

First, you can configure Core Dump on each ESXi hosts on SSH.
Connect to your ESXi host from SSH. Change port configuration if necessary.

# esxcli system coredump network set --server-ip=IpvCenterServer --server-port=6500 --interface-name=vmk0
# esxcli system coredump network set --enable true

You can verify your configuration with this command.

# esxcli system coredump network get

Check if your configuration works fine.

# esxcli system coredump network check

Second, you can use Host profiles to apply same the configuration on all of your ESXi hosts.
Edit this value only and apply of all of your ESXi hosts in maintenance mode.

Validated your configuration

You can verify your configuration, but you need a crash on of your ESXi Host. You can use command vsish to force failed one of your ESXi host. Please look at a previous post http://blog.purplescreen.fr/?p=1556.
All crashes are stored on /var/core/netdumps. You can list all crashes on your vCenter server.

# ls -lha /var/core/netdumps