upgrading an old host from 151038 to 151046. It mostly worked. However, it got really pissy about the "UsePAM no" command in sshd_config. <sigh>

other than that it seems to have survived.

oh, it's also upset about svc:/system/hal

I think it warns about UsePam as deprecated, but did it do something else?

it failed to start ... several hundred times.

"Start method failed repeatedly, last exited with status 255"

I removed the line from sshd_config and it's starting now.

though it's ignoring my duo 2FA so I'll need to fix that.

pam, as far as know, has always been permanently on in the omnios openssh. I didn't think it was fatal to have it in the configuration file though, I wonder when that happened?

it also takes for-freakin-ever to mount filesystems after boot. I'm waiting to see if anything more shows up on the console about failed starts.

it was set to 'no' for this host for reasons long lost to history.

ah, it's failing to start svc:/system/hal so that's what's causing slow mounts.

STATE STIME FMRI

maintenance 13:20:01 svc:/system/hal:default

Reason: Start method exited with $SMF_EXIT_ERR_FATAL.

well that's useful information... not.

[ Jun 7 13:15:50 Executing start method ("/lib/svc/method/svc-hal start"). ]

hal failed to start: error 2

[ Jun 7 13:20:01 Method "start" exited with status 95. ]

but it starts fine with svcadm clear and svcadm enable.

so it's something it doesn't like during boot.

error 2 seems to be that one of its children timed out

- more than 250s for device probing

ah.. interesting.

I have to step afk, but good luck getting to the bottom of it, and welcome to r46!

thanks... I think :)

"dmesg | grep -i warning" has much to say about error recovery for port 4 and 5.