[BlueOnyx:16166] Re: Scientific Linux 7.0 released

[Michael Stauber]

Hi Ernie,

> The thing that gets me is there seems to be no transition period, every
> hosting control panel suddenly has to have a complete rewrite on process
> control. I have seen on the debian Wiki they have a package for systemd 
> which creates the old init.d/ scripts structure to maintain some compatibility, 
> perhaps there is something like that for E7?

Yeah, they have some kludges set up which kinda work like this:

[root at 5209r /]# /sbin/service httpd restart
Redirecting to /bin/systemctl restart  httpd.service

It doesn't work reliably, though, as the entire SysV init stuff is
missing which would prevent the same service from being started more
than once. For some services this kludge works more or less OK, for
others it doesn't.

Systemd has it's own mechanism for this, which is radically different.
Ideally each daemon needs it's own Systemd config file, like each SysV
init script needs its RC start/stop script.

The really sad part is this: Systemd might be great for laptops,
desktops and workstations. Faster boot times, parallel init of things
and such. But for servers that isn't an issue and offers just drawbacks
and complexity where you simply could do without that.

Unfortunately RedHat went entirely for Systemd and most services no
longer have the SysV init scripts. So there is no way to change back to
SysV without breaking the ability to install updates.

You know what the most common smack down reply is when someone has a
complaint about Systemd? It's "Deal with it." It's like the inmates have
taken over the asylum. This "my way or the highway" attitude just sparks
my rebellious traits to not give in to that kind of hostage taking.
Sadly there are just very few alternatives and none of them has me
excited either.

We'll see. But like said: There are some other fundamentals that need
sorting first.

-- 
With best regards

Michael Stauber