Re: SysVinit problem?

2000-09-03 Thread Henrique M Holschuh
On Sun, 03 Sep 2000, Bruce Richardson wrote:
 All the user runlevel directories, /etc/rc1.d/ through to /etc/rc5.d/,
 have exactly the same contents and they're all start scripts, no kill
 scripts.  If I telinit from (for example) runlevel 2 to 4, nothing
 happens except for the sending a term/kill signal to all processes
 message.  The console I type it at stays the same but all the other ttys
 freeze until I telinit back to the original runlevel.

This is ok, Debian doesn't use runlevels 3-5 for anything by default AFAIK,
and they're mostly equal to runlevel 2 (I think /etc/inittab has some stuff
which is different, simply to show it can do that).

BTW, there's an utterly braindamaged behaviour in many (most?) daemon
packages during upgrade: They will start their daemons regardless of the
current runlevel, so keep this in mind during upgrades if you hand-trimmed
your runlevels to actually mean something.

Proposing a fix to this is in my TODO list. The code is rather easy, really,
but requires a policy change as almost all packages who have something in
/etc/init.d will have to be fixed.

-- 
  One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie. -- The Silicon Valley Tarot
  Henrique Holschuh


pgpBaBhEhwKdm.pgp
Description: PGP signature


Re: SysVinit problem?

2000-09-03 Thread Bruce Richardson
On Sun, Sep 03, 2000 at 10:12:33AM -0300, Henrique M Holschuh wrote:
 
 This is ok, Debian doesn't use runlevels 3-5 for anything by default AFAIK,
 and they're mostly equal to runlevel 2 (I think /etc/inittab has some stuff
 which is different, simply to show it can do that).

I thought that might be the case but I'm still concerned about the
freezing ttys.  I can't believe that it's intended behaviour.

 
 BTW, there's an utterly braindamaged behaviour in many (most?) daemon
 packages during upgrade: They will start their daemons regardless of the
 current runlevel, so keep this in mind during upgrades if you hand-trimmed
 your runlevels to actually mean something.

Hmmm.  A newly installed package won't know which levels you want it
running in but one being upgrade ought to be able to check.

 
 Proposing a fix to this is in my TODO list. The code is rather easy, really,
 but requires a policy change as almost all packages who have something in
 /etc/init.d will have to be fixed.

I'm new to Debian and only just finished reading the policy docs atc.
I suppose some extended version of update-rc.d is the thing for that.




-- 
Bruce

The good news is that in 1995 we will have a good operating system and
programming language; the bad news is that they will be Unix and C++.
-- Richard P. Gabriel



Re: SysVinit problem?

2000-09-03 Thread Henrique M Holschuh
On Sun, 03 Sep 2000, Bruce Richardson wrote:
 On Sun, Sep 03, 2000 at 10:12:33AM -0300, Henrique M Holschuh wrote:
 I thought that might be the case but I'm still concerned about the
 freezing ttys.  I can't believe that it's intended behaviour.

It is not, but it may be either something weird in /etc/inittab (I seem to
recall some ttys aren't restarted in all runlevels by default. They WILL
freeze), or your tty driver is bonkers (I recommend fbgetty for consoles,
mgetty for serial lines).

  BTW, there's an utterly braindamaged behaviour in many (most?) daemon
  packages during upgrade: They will start their daemons regardless of the
  current runlevel, so keep this in mind during upgrades if you hand-trimmed
  your runlevels to actually mean something.
 
 Hmmm.  A newly installed package won't know which levels you want it
 running in but one being upgrade ought to be able to check.

My point exactly.

  Proposing a fix to this is in my TODO list. The code is rather easy, really,
  but requires a policy change as almost all packages who have something in
  /etc/init.d will have to be fixed.
 
 I'm new to Debian and only just finished reading the policy docs atc.
 I suppose some extended version of update-rc.d is the thing for that.

Yes, and the new script would be provided by the file-rc and sysvinit
packages (and any other future /etc/init.d wrappers/handlers).

-- 
  One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie. -- The Silicon Valley Tarot
  Henrique Holschuh


pgpaGmSo47SmQ.pgp
Description: PGP signature


Re: SysVinit problem?

2000-09-03 Thread Bruce Richardson
On Sun, Sep 03, 2000 at 01:42:17PM -0300, Henrique M Holschuh wrote:
 On Sun, 03 Sep 2000, Bruce Richardson wrote:
  On Sun, Sep 03, 2000 at 10:12:33AM -0300, Henrique M Holschuh wrote:
  I thought that might be the case but I'm still concerned about the
  freezing ttys.  I can't believe that it's intended behaviour.
 
 It is not, but it may be either something weird in /etc/inittab (I seem to
 recall some ttys aren't restarted in all runlevels by default. They WILL
 freeze)

Aye, that is the case.  Only tty1 is restarted in all 5 levels, the
others only in 2 or 3.  This seems bizarre and arbitrary, seeing that
otherwise the runlevels are identical.

-- 
Bruce

If the universe were simple enough to be understood, we would be too
simple to understand it.



Re: SysVinit problem?

2000-09-03 Thread W. Paul Mills
Bruce Richardson [EMAIL PROTECTED] wrote:
: On Sun, Sep 03, 2000 at 10:12:33AM -0300, Henrique M Holschuh wrote:
: 
: This is ok, Debian doesn't use runlevels 3-5 for anything by default AFAIK,
: and they're mostly equal to runlevel 2 (I think /etc/inittab has some stuff
: which is different, simply to show it can do that).

: I thought that might be the case but I'm still concerned about the
: freezing ttys.  I can't believe that it's intended behaviour.


Look at /etc/inittab. Find the section on getty's. There is your
answer. Only one getty is run above runlevel 3.



-- 
*** Running Debian Linux ***
*   For God so loved the world that He gave his only begotten Son,  *
*   that whoever believes in Him should not perish...John 3:16  *
* W. Paul Mills  *  Topeka, Kansas, U.S.A.  *
* EMAIL= [EMAIL PROTECTED]  *  WWW= http://Mills-USA.com/  *
* Bill, I was there several years ago, why would I want to go back? *
* pgp public key on keyservers everywhere? */
--