here is an update to this, just for others if trying, and a couple of questions as well - I know this is way outside the scope of support…
it does appear I made my machine init scripts not run properly on shutdown, even a -f does not bring it down (it kills ssh and apache so I have no insight after that)… After 10 minutes of ping replies never ending, I issued a reboot from the dh panel, I dont know what that process does, I assume it sends power-button signal or something, and perhaps a hard reset if no response within a certain time, who knows (?), it takes about 8 minutes from a panel-reboot command till it actually reboots.
so, couple of questions for anyone who might know;
is there any user accessible console anywhere?
other vps hosts I deal with usually have some sort of interface for this, wether vnc based or redirected console stream.
is there any options for more granular control of the vm?
such as sending a power-signal, vm-reset or power-off/on options
other then the dreamhost entries for sources list, is the vm os based on some certain vserver distro or something?
so, just some info for others who might attempt this… Some of the issues I had are listed below.
This installation really seems heavily customized, like why wouldn’t there be a /dev/initctl ? it has a lot of custom packages, it might be related to security updates and such, but it kind of sucks when some core packages are custom… itwas a lot of work and hairpulling, I dont know why but I had quite the few issues getting the packages aligned to debian standard, I found that doing a purge on all the dh pakages was the quickest way to get started, removed all the dh sources, added debian ones and updated, removed the dh entries first, then went through the rest of the obsolete/local ones… I finally got it all down to a bare standard with apache and php, had some issues here and there, had to manually delete quite a few things which the packages didnt do cleanly, such as /etc/perl - and I did --reinstall on several core items, probably didnt need them all but doing that on all the packages I could think of related to the current issue got it solved in most cases. Some manual fumbling on package selection to get rid of courier (change MTA).
/dev/initctl did not exist, I did some --reinstall of init, base, and sysv packages, the node was created, but still there are issues, a reboot command shows this:
shutdown: timeout opening/writing control channel /dev/initctl
init: timeout opening/writing control channel /dev/initctlthen returns to shell,
a reboot -f, like mentioned, kills ssh and apache, and probably a ton of other stuff, but still wont reboot… for now I will just have to do a -f and then wait 8 minutes for the panel reboot to finish.
some day I’ll get back to figuring out the init scripts - a bit tough to work on without a console available.
During my cleanup I had the sources list down to this
deb http://some.mirror.dom/debian/ squeeze main contrib non-free
deb http://security.debian.org/ squeeze/updates main contrib non-free
Then I added this after the cleanup was done, for some availability
deb http://www.backports.org/debian/ squeeze-backports main contrib non-free
I also went in and set my apache start-servers and min/max spare servers to very low numbers - as my vm is one with little memory.
I am not all done cleaning out yet, but the system is functional (except the shutdown issue), if anyone wants a copy of the dpkg --get-selections they are at http://dh.stokkeland.net/getsel for now