dhcpd: no such user: dhcpd

[mircsicz]

I've imported an existing config into a fresh nanosd 16.1.8 CF Install on an ALIX.

And I only had to do so because the system crashed while trying to upgrade with disabled RAM-Disks...

On the first but after restore DHCPD was running and I got an IP, a few hours later dhcpd seems no to be running and when trying to start it I see the following log:

Code: [Select]

May 22 12:00:26 dhcpd: no such user: dhcpd
May 22 12:00:26 dhcpd: For info, please visit https://www.isc.org/software/dhcp/
May 22 12:00:26 dhcpd: All rights reserved.
May 22 12:00:26 dhcpd: Copyright 2004-2016 Internet Systems Consortium.
May 22 12:00:26 dhcpd: Internet Systems Consortium DHCP Server 4.3.3-P1
May 22 12:00:21 dhcpd: exiting.
May 22 12:00:21 dhcpd:
May 22 12:00:21 dhcpd: process and the information we find helpful for debugging..
May 22 12:00:21 dhcpd: before submitting a bug. These pages explain the proper
May 22 12:00:21 dhcpd: bugs on either our web page at www.isc.org or in the README file
May 22 12:00:21 dhcpd: than a configuration issue please read the section on submitting

So I ran adduser on the console and could successfully start dhcpd from the webif afterwards...

If your're willing to debug I can supply the config!

[franco]

This used to happen sparsely, fixed with 16.1.13:

https://github.com/opnsense/changelog/blob/master/doc/16.1.13#L28

To manually fix... Can you first let me know what these commands yield?

# ls -lah /var/db
# ls -lah /root/var/db/pkg

[mircsicz]

Hi Franco,

thx for all your effort! ;-)

Code: [Select]

root@router:~ #  ls -lah /var/db
total 36
drwxr-xr-x   7 root    wheel   432B May 23 05:44 .
drwxr-xr-x  13 root    wheel   396B May 21 14:45 ..
-rw-r--r--   1 root    wheel     0B May 21 14:51 currentipsecpinghosts
drwxr-xr-x   2 root    wheel   108B May 21 14:48 dnscache
drwxr-xr-x   2 root    wheel     0B May 21 14:43 entropy
-rw-r-----   1 root    wheel    62B May 21 14:44 growlnotices_lastmsg.txt
-rw-r-----   1 root    wheel     0B May 22 14:44 ipsecpinghosts
-rw-r-----   1 root    wheel     8B May 23 05:44 ntpd.drift
drwxr-xr-x   2 root    wheel     0B May 21 14:51 pingmsstatus
drwxr-xr-x   2 root    wheel     0B May 21 14:51 pingstatus
lrwxr-xr-x   1 root    wheel    16B May 21 14:43 pkg -> /root/var/db/pkg
drwxr-xr-x   2 nobody  wheel   720B May 21 14:45 rrd
-rw-r-----   1 root    wheel    12B May 22 14:44 wan_cacheip
-rw-r-----   1 root    wheel    12B May 22 14:44 wan_ip
root@router:~ # ls -lah /root/var/db/pkg
total 5321
drwxr-xr-x  2 root  wheel   512B May 21 14:30 .
drwxr-xr-x  3 root  wheel   512B May 21 14:07 ..
-rw-r--r--  1 root  wheel     0B May 21 14:30 OPNsense.meta
-rw-r--r--  1 root  wheel   4.7M May 21 14:30 local.sqlite
-rw-r--r--  1 root  wheel   2.5K May 21 14:30 local.sqlite-journal
-rw-r--r--  1 root  wheel   502K May 21 14:30 repo-OPNsense.sqlite


[franco]

That looks ok now, one more to be sure:

# pkg info

It should show all packages...

If that is the case updating to 16.1.14 would likely fix this for you if the dhcpd issue persists on reboot.

[mircsicz]

Hi franco,

as I'm on ALIX I can't upgrade....

and "pkg info" just returns nothing

But it's reproducable, just changed the CF again, and restored the config and the problem was back!

[interfaSys]

This was still happening on 16.1.14, especially after a kernel panic+reboot

[franco]

That means we'll have to start backing up the package database because we really can't trust UFS at all.

[interfaSys]

I just wanted to add that in my case, restarting the box solved the problem, unless this happens at random of course...

[franco]

The first reboot may not recover completely, but the second one does... we've seen this, but don't really know why.

I've added pkg database dump/restore though I do not think it will help very much.

https://github.com/opnsense/core/commit/a6132af6e4d8d904a6b40770e7ef9bce9918584e

mircsicz, in your case the database is already gone. You can force an upgrade for all packages like this:

# pkg install opnsense

That should bring it back.