Managing Core Dumps

With esxcli system coredump, you can manage local diagnostic partitions or set up core dump on a remote server in conjunction with ESXi Dump Collector. For information about ESXi Dump Collector, see the vSphere Networking documentation.

Managing Local Core Dumps with ESXCLI

The following example scenario changes the local diagnostic partition with ESXCLI. Specify one of the connection options listed in Connection Options for vCLI Host Management Commands in place of <conn_options>.

To manage a local diagnostic partition

1	Show the diagnostic partition the VMkernel uses and display information about all partitions that can be used as diagnostic partitions.

esxcli <conn_options> system coredump partition list

2	Deactivate the current diagnostic partition.

esxcli <conn_options> system coredump partition set --unconfigure

The ESXi system is now without a diagnostic partition, and you must immediately set a new one.

3	Set the active partition to naa.<naa_ID>.

esxcli <conn_options> system coredump partition set --partition=naa.<naa_ID>

4	List partitions again to verify that a diagnostic partition is set.

esxcli <conn_options> system coredump partition list

If a diagnostic partition is set, the command displays information about it. Otherwise, the command shows that no partition is activated and configured.

Managing Core Dumps with ESXi Dump Collector

By default, a core dump is saved to the local disk. You can use ESXi Dump Collector to keep core dumps on a network server for use during debugging. ESXi Dump Collector is especially useful for Auto Deploy, but supported for any ESXi 5.0 and later host. ESXi Dump Collector supports other customization, including sending core dumps to the local disk.

ESXi Dump Collector is included with the vCenter Server autorun.exe application. You can install ESXi Dump Collector on the same system as the vCenter Server service or on a different Windows or Linux machine. See vSphere Networking.

You can configure ESXi hosts to use ESXi Dump Collector by using the Host Profiles interface of the vSphere Web Client, or by using ESXCLI. Specify one of the connection options listed in Connection Options for vCLI Host Management Commands in place of <conn_options>.

To manage core dumps with ESXi Dump Collector

1	Set up an ESXi system to use ESXi Dump Collector by running esxcli system coredump.

esxcli <conn_options> system coredump network set --interface-name vmk0 --server-ipv4=1-XX.XXX --port=6500

You must specify a VMkernel port with --interface-name, and the IP address and port of the server to send the core dumps to. If you configure an ESXi system that is running inside a virtual machine, you must choose a VMkernel port that is in promiscuous mode.

2	Enable ESXi Dump Collector.

esxcli <conn_options> system coredump network set --enable=true

3	(Optional) Check that ESXi Dump Collector is configured correctly.

esxcli <conn_options> system coredump network get

The host on which you have set up ESXi Dump Collector sends core dumps to the specified server by using the specified VMkernel NIC and optional port.

Managing Core Dumps with vicfg-dumppart

The following example scenario changes the diagnostic partition. Specify one of the connection options listed in Connection Options for vCLI Host Management Commands in place of <conn_options>.

To manage a diagnostic partition

1	Show the diagnostic partition the VMkernel uses.

vicfg-dumppart <conn_options> -t

2	Display information about all partitions that can be used as diagnostic partitions. Use -l to list all diagnostic partitions, -f to list all diagnostic partitions in order of priority.

vicfg-dumppart <conn_options> -f

The output might appear as follows.

Partition name on vml.mpx.vmhba36:C0:T0:L0:7 -> mpx.vmhba36:C0:T0:L0:7

3	Deactivate the diagnostic partition.

vicfg-dumppart <conn_options> -d

The ESXi system is now without a diagnostic partition, and you must immediately set a new one.

4	Set the active partition to naa.<naa_ID>.

vicfg-dumppart <conn_options> -s naa.<naa_ID>

5	Run vicfg-dumppart -t again to verify that a diagnostic partition is set.

vicfg-dumppart <conn_options> -t

If a diagnostic partition is set, the command displays information about it. Otherwise, the command informs you that no partition is set.