any problem going from 9.x (don't laugh) to 11 directly?

Tim Daneliuk tundra at
Thu Feb 15 00:05:07 UTC 2018

On 02/14/2018 05:51 PM, Tim Daneliuk wrote:
> On 02/14/2018 05:29 PM, Randal L. Schwartz wrote:
>> Doug> There are enough library changes that most services just will not
>> Doug> start properly and tend to hang.  You are left with a system that
>> Doug> is non-responsive.  The one approach I used successfully one time
>> Doug> was to comment virtually everything other than sshd out in
>> Doug> rc.conf.  You have to make sure you have really killed sendmail
>> Doug> though.  Setting it to NO is not enough.  Otherwise it will
>> Doug> eventually give up and give you back service, but it takes a long
>> Doug> time.
>> Great advice!  Thanks.
> Another strategy might be to put a second drive in the system and install
> a clean version of 11 and all the needed packages/ports on that.  Then -
> since it can mount your existing 9.x disk, incrementally copy over your
> data and custom configurations.  In my experience, this turns out to be
> a lot quicker and cleaner that trying to do in-place upgrades.  Moreover,
> when you are all done, you still have your original, untouched disk.
> Disks are cheap :)

Oh, another way I've done this in inside VMs.  Once configured and built
to my satisfaction, I just tarball off the entire contents of the VM's filesystem.
Then, I pour it onto a bare "real" drive that's been partitioned and formatted

On Linux systems, I do some this with docker instances, although full OS imaging
is a good use case for this approach.

Some notes I've jotted down over the years:

Tim Daneliuk     tundra at
PGP Key:

More information about the freebsd-questions mailing list