What can QA do productive when the AUT is down/unavailable?

I could think of the following - - look at splunk logs to see the cause - learn about any deployment issues - test negative scenarios which can validate if these are working as expected - write blueprint code/scripts for automation etc.

is there anything else that can be done?