Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 20 Jan 2021 18:54:31 +0000
From:      Colin Percival <cperciva@tarsnap.com>
To:        Rafal Lukawiecki <raf@rafal.net>
Cc:        freebsd-cloud@freebsd.org
Subject:   Re: FreeBSD on AWS Graviton (t4g)
Message-ID:  <010001772125f524-74bf0b19-cd7c-4091-b0bb-e9fcb9ededdf-000000@email.amazonses.com>
In-Reply-To: <30C35A29-17E2-481D-A335-5B1FBB9FB22C@rafal.net>
References:  <01000176816d1718-26a08928-e08c-466a-8550-552bc746d81f-000000@email.amazonses.com> <C4D2ACA9-BFFE-49C1-B8AA-72E32C9DB6C9@rafal.net> <30C35A29-17E2-481D-A335-5B1FBB9FB22C@rafal.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On 1/20/21 5:12 AM, Rafal Lukawiecki wrote:
> Colin, I am only slowly realising the impact of the points you have made. :)
> 
>>> On 20 Dec 2020, at 18:33, Colin Percival <cperciva@tarsnap.com
>>> <mailto:cperciva@tarsnap.com>> wrote:
>>>
>>> * The EC2 API "shutdown" / "reboot" message is sent via a pl061 GPIO device
>>> and we're dropping it on the floor; consequently "reboot" does nothing at
>>> all, while "shutdown" does nothing until EC2 times out and does a hard
>>> poweroff (without filesystems being unmounted etc).
> 
> I have noticed a bunch of artefacts left over from unclean shutdowns, which I
> assume are caused by the above not calling the shutdown rc.d scripts. Do you
> know if this is likely to get implemented any time soon, or are you aware of a
> simple workaround?

If you run 'shutdown' from inside the EC2 instance it should work fine.  Just
don't use the EC2 API to stop the ARM instances.

-- 
Colin Percival
Security Officer Emeritus, FreeBSD | The power to serve
Founder, Tarsnap | www.tarsnap.com | Online backups for the truly paranoid



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?010001772125f524-74bf0b19-cd7c-4091-b0bb-e9fcb9ededdf-000000>