Timing out on graceful shutdown during Terraform commands

[findkim]

There's an underlying issue somewhere causing the graceful shutdown attempt to timeout after 10 seconds when the shutdown occurs during a Terraform command.

^C2020/09/25 00:57:23.073189 [INFO] signal received to initiate graceful shutdown: interrupt
2020/09/25 00:57:23.075615 [ERR] (client.terraformcli) unable to change workspace: "print"
2020/09/25 00:57:23.075638 [INFO] (task) stopping retry of 'Init print'
2020/09/25 00:57:23.075641 [ERR] (driver.terraform) init (skip apply) for 'print'
2020/09/25 00:57:23.075669 [ERR] (cli) error running controller: [ERR] (controller.readwrite): could not apply: Error tf-init for 'print': context canceled
2020/09/25 00:57:33.078077 [INFO] graceful shutdown timed out, exiting

It's happening at some frequency during the e2e test as well (1, 2).

--- FAIL: TestE2EBasic (21.06s)
    e2e_test.go:53: 
        	Error Trace:	e2e_test.go:53
        	Error:      	Received unexpected error:
        	            	exit status 12
        	Test:       	TestE2EBasic

Would be useful to improve shutdown during a Terraform command