cannot open cd0a:/etc/random.seed: No such file or directory

[Amithapr]

Hi,

I am planning to upgrade my OpenBSD version to the OpenBSD 6.1 latest version. When I try to install OpenBSD gave the error "cannot open cd0a:/etc/random.seed: No such file or directory"

After the installation, OpenBSD did not boot and triggered the error " cannot open hd0a:/etc/random.seed: No such file or directory "

After searching the internet, I just enabled the console output to com0 with the default port values, then the above problem solved. But since the output is directed to console as it should be, it's not displaying anything during the boot process

I can access my OpenBSD machine via ssh. Can I get this thing sorted out so the above error is not triggered and the console output is displayed to the monitor?

Thanks.

[jggimi]

The "missing random.seed" message is unrelated to any installation or upgrade problems you may be having. The file is created/written by rc(8) during shutdown, and then read by rc() again during boot. It is one of many entropy feeds to arc4random(4). It is always missing on a new installation, because there was no prior shutdown to create it.

https://youtu.be/aWmLWx8ut20 (video)
http://www.openbsd.org/papers/hackfe...dom/index.html (slides)

[ocicat]

Originally Posted by Amithapr

I am planning to upgrade my OpenBSD version to the OpenBSD 6.1 latest version.

...and what is the version of OpenBSD you are beginning with?

[Amithapr]

Hi jggimi & ocicat,

Sorry for the late reply. Thanks for the information Jggimi. I will try these videos and do a fresh installation.

Thanks for your reply Ocicat, I'm upgrading my firewall from OpenBSD 5.3 to OpenBSD 6.1

Thanks & Regards,

Amitha

[ocicat]

Originally Posted by Amithapr

I'm upgrading my firewall from OpenBSD 5.3 to OpenBSD 6.1

The OpenBSD project only verifies the upgrade process through subsequent versions: 5.3 to 5.4, 5.4 to 5.5, 5.5 to 5.6, ..., all the way to 6.0 to 6.1. Skipping versions is not supported.

Unless you have lots of time & patience, it would be significantly quicker to simply reinstall directly to OpenBSD 6.1. Of course, you would need to backup all important files first.

[Amithapr]

Thanks for the information Ocicat. Yes I'm planning to do a fresh installation, sorry for the misleading word "upgrade".
I know this is bit difficult, since I have to do lot of learing to get the system up and running since so many new features and methods are in the 6.1 version.

Thanks

[jggimi]

Originally Posted by Amithapr

After the installation, OpenBSD did not boot ... I can access my OpenBSD machine via ssh.

If you can access it via SSH, it booted. So one of these two statements is incorrect, or, you misstated something.

As always, Amitha, we can only see the information you decide to provide.

If you are able to access a 6.1 installation via SSH, providing a dmesg will tell us what the OS knows about your hardware.

[Amithapr]

Hi Jggimi,

Yes I could login to the OpenBSD firewall via ssh only after redirecting the console output to com0. If you don't do that as I mentioned in my starting thread those errors triggers without booting the OS.

Yesterday I did a fresh installation after referring to Ociact's information. I did not directed the console output to com0, the only changes that I did was accepting the automatic partition layout and when it prompts to reboot after the installation I did not removed the CD from the cd player.

It rebooted, then it booted from cd again and prompted for installation options. Then I selected the shell option, from shell I rebooted the machine for the second time and removed the CD. Finally it booted properly.

Thanks

[Amithapr]

Hi,

Finally I fount the mistake that it would have been. I tried to create a custom partition layout several times it triggers this error. Then I tried the default partition layout for i386, THE ERROR DISAPPEARED.

CONCLUTION, if i'm right: You have to follow the partition layot for the architecture but the partition sizes you can change

Thanks

[jggimi]

At OpenBSD 6.1-release, the OpenBSD install/upgrade scripts will create the random.seed file in /mnt/etc. /mnt is where the new root partition is mounted.

dd if=/dev/random of=/mnt/etc/random.seed bs=512 count=1 status=none

On a running system that is shut down, init(8) calls rc(8) to run its shutdown routines, which includes a call to the random.seed() function to apply the old seed and create a new one.

random_seed() {
        # push the old seed into the kernel
        dd if=/var/db/host.random of=/dev/random bs=65536 count=1 status=none
        chmod 600 /var/db/host.random
        # ... and create a future seed
        dd if=/dev/random of=/var/db/host.random bs=65536 count=1 status=none
        # and create a seed file for the boot-loader
        dd if=/dev/random of=/etc/random.seed bs=512 count=1 status=none
        chmod 600 /etc/random.seed
}

[jggimi]

Thinking about this, if /etc is not included within the root partition, the bootloader will not find the /etc/random.seed file, and you will see this message. The bootloader reads the root partition for its configuration file (/etc/boot.conf), kernels, and the random seed file (/etc/random.seed).

TL;DR: do not deploy /etc in its own partition.

[Amithapr]

Hi Jggimi,

Thanks for the information.