Shutdown Ironport ESA

[edisonguerrero]

Hi comunity,

I have a quiestion about the shutting down of an Ironport ESA C370. In a CLI session, i write the "shutdown" command. After thar, the message "Enter the number of seconds to wait before forcibly closing connections. [30]>" appears. I have put different values (since 0 to 200 ) and the message "System shutting down. Please wait while the queue is being closed....." appears and the CLI session is closed with the message "The system will power off automatically."

But the appliance never shuttdown. The ping is allways working and if I try a new cli or gui session, it works normally as if the shutting down command never were wrote.

Any advice about it?

[Libin Varghese]

Would need more review. When issuing a shutdown command on the appliance all background services are stopped before the appliance shuts down. Any service not shutting down correctly can prevent the process from completing.

You can open a TAC case to get that checked.

Regards,

Libin Varghese

[AsepRizkiKhoeisman1481]

Hi Libin,

we had a cluster of ESA C300V and we had a plan to shutdown them for maintenance purposes. how is the proper step to shutdown ESA on the cluster mode? should I disjoin the cluster and manually shutdown them one by one?

Regards,

Asep Rizki Khoerisman

[Ken Stieers]

You do not need to disjoin them from the cluster. On each ESA suspend all of the listeners and wait for the workqueue to empty, then shut it down.

Do your maintenance, boot them up. Once you can login to he gui, resume the listeners.

[Libin Varghese]

You can use  the 'shutdown' command from the CLI, or the Shutdown/Reboot  option listed under the system administration tab in the GUI of each appliance.

It's always recommended that all virtual ESA's be shutdown before any maintenance activity on the VM server or platform to avoid chances of queue corruption or database corruption.

Disconnecting an appliance from the cluster is not a mandatory step for device shutdown.

To elaborate, if one machine in a cluster is shut down while others are still powered on, you'll see alerts being generated stating the shutdown machine could not be reached since it performs a keep-alive with all machines every 5 minutes or so.

If you wish to avoid alerts from being generated, you can certainly administratively disconnect appliances from the cluster (using clusterconfig -> disconnect) from any one of the machines and reconnect them after the maintenance.

However, the alerts themselves would not hinder the shutdown being performed.

Regards,

Libin