Re: how to make Debian less fragile (long and philosophical)
Wednesday, August 18, 1999, 2:45:34 PM, Justin wrote:
> I have not run 10,000 machines over my career, and I have seen more
> than one failure. I have run probably not even 50 machines, and
> five or six times I've needed statics. In my experience, then, the
> ratio is closer to 1 in 10.
Oh, the lovely personal experience. At one site I was an admin at, on
FreeBSD machines, I had one machine get to a point where I could have used
statics. However, IIRC in this thread someone pointed out that BSD (I assume
that includes Free) *does* have statics. Didn't much help, I had to boot from
a rescue floppy.
That was out of a good 20 machines that were running at that moment.
Maybe a total of 50 during my tenure there. I'm now at a location with many,
MANY more machines. A few hundred at least. I cannot think of a single
report of a failure that needed statics or where statics would have helped.
Of course, we also run in an N+1 architecture so if a machine does fail we
just pull it out of the farm, replace it with another machine ready to go and
service is never interrupted.
So, your experience, which you say is <50 has have 5-6 times. Mine with
>100 on the conservative side, has had 0 where statics would have helped.
I bet I could find a lot more people who have 0 and you that have 5-6.
Personal experience, btw, means dick.
> If you want to try and analyze this a different way, estimate what in
> your opinion is the mean time to failure for a Debian machine. How
> many days do you think would pass, with normal upgrades and normal
> operation, before some kind of error, hack, mistake, or bug would
> bring your machine down such that you needed statics.
In the two years across the three machines I do have Debian on I've had
one case where I *needed* to reboot to fix it. A static would not have
helped. It was the syslogd/klogd problem on unstable that prevented root
logins completely. Had to reboot to single user mode to get root to fix.
Statics would not have helped.
In fact, I have sash on all my Debian machines and I've yet to have a need
for it.
> What I intended to show you was that even if Debian is pretty reliable,
> it is hard to be so reliable that good, durable recovery tools are
> not still important.
N+1 makes it pretty much unimportant. That is something Debian cannot
control and something that anyone who wants as close to 100% uptime on
services as possible should be doing.
> I think that statics should be the default, that every user should get
> for their own good. Intelligent users, who are damn sure what they
> are doing, and know that their system is not important, should have
> the option of not installing it--but that shouldn't be the default.
As I pointed out several times now, they do. Anyone who is in a position
of needing such reliability should also have the knowledge and skill to
compile their own. In fact, they should have the skills to compile their own,
make packages of it to their specifications, and use those packages in their
environment. They should also have the intelligence and backing to have an
N+1 environment so they don't have to worry about a single point of failure on
any machine when it comes to services.
--
Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
ICQ: 5107343 | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------
Reply to: