Old Possum's book of practical hosts.

[Menno]

Engfeh.

Just spent these last two days renaming all of the hosts in my nice
test/development domain. InconCEIvable how many places that hostname
pops up in. They were originally named after cartoon characters, but
it was deemed that someone might see it and experience a brief moment
of glee, which is clearly inappropriate.

So now we named the buggers after Artists: Leonardo, Donatello,
Raphael, Michelangelo and the famous 14th century Dutch renaissance
painter Hans Splinter. And thus is my brief moment of glee preserved.

What is it about hostnames that so gets up people's snouts? I've seen
horrid constructions. There's the school of thought that thinks it
would be a good idea to stuff as much information as possible into the
hostname, leading to names like "rs6000-aix51-2002-03-01-1". (Which of
course runs AIX 5.3, because renaming it after the upgrade would have
been Difficult). Then, there is the "We need to know that it belongs
to Our Company" school of thought, leading to
MyCompany0001.mycompany.com. Or the "Lack of imagination" school of
thought: "box0001".

And then, with the naming convention well and truly established, the
DNS servers are duly buggered up, so that everybody uses the IP
addresses anyway.

Anyways, everything is working again. Time for beer and merriment.

[Roger Burton West]

Post by Menno
What is it about hostnames that so gets up people's snouts? I've seen
horrid constructions. There's the school of thought that thinks it
would be a good idea to stuff as much information as possible into the
hostname, leading to names like "rs6000-aix51-2002-03-01-1". (Which of
course runs AIX 5.3, because renaming it after the upgrade would have
been Difficult). Then, there is the "We need to know that it belongs
to Our Company" school of thought, leading to
MyCompany0001.mycompany.com. Or the "Lack of imagination" school of
thought: "box0001".

I think that much of this flows from two principles:

(1) We don't understand hierarchical namespace. (Obvious in the
.com/.net/.org land-grab but widely applicable elsewhere too.)

(2) We aren't aware of the concept of aliases; a host must have only one
name.

Though, as you point out, there is a strong undercurrent of "no personal
touches" in many companies.

The boxes I've built for the legal charity for which I work (hostnames
are not relevant to outside use) are learned-hand, mansfield, diplock,
bruneis, dworkin, williams, ...

Post by Menno
And then, with the naming convention well and truly established, the
DNS servers are duly buggered up, so that everybody uses the IP
addresses anyway.

The standard things I do when I'm called on to "fix this office
network":

- vafgnyy jbexvat QAF, naq chfu vg guebhtu QUPC;
- vafgnyy jbexvat AGC;
- vafgnyy n JVAF freire gb phg qbja ba nyy gung oebnqpnfg genssvp.

R

[Menno]

On Fri, 23 Jun 2006 17:16:12 UTC, Roger Burton West

Post by Roger Burton West
The standard things I do when I'm called on to "fix this office
- vafgnyy jbexvat QAF, naq chfu vg guebhtu QUPC;

Yup. Did that yesterday. I took the opportunity to re-install certain
servers that, due to some ill-advised maneuvers, had some of their
entrails hanging out.

Post by Roger Burton West
- vafgnyy jbexvat AGC;

Did that today (as an afterthought).

Post by Roger Burton West
- vafgnyy n JVAF freire gb phg qbja ba nyy gung oebnqpnfg genssvp.

You know, Installment 1 of this net had everything. I thought I'd be
nice to my friends and colleagues and install a Fnzon domain, complete
with a nice controller, home directories, Roman Profiles, file
services and whatnot. (The Fnzon guide refers to it as "making users
happy", and who wouldn't want happy users?) So when the time came for
my Windows-using friends to enter the realm of Separate Test Networks,
the first thing they did was put in a PDC. So this time round, I
thought: "Buggre This For An Game Of Soldiers" and just stuck in YQNC
for the NVK and Yvahk boxen. No JVAF for you.

[Brian Kantor]

Post by Menno
What is it about hostnames that so gets up people's snouts?
And then, with the naming convention well and truly established, the
DNS servers are duly buggered up, so that everybody uses the IP
addresses anyway.

And then the auditor comes in and insists that the hostnames must be
kept secret (i.e., not in the hosts table) because they might serve
to direct an attacker to a weak host, so you now have to encrypt
the hosts table. Of course, said auditor received his computer security
from a book published before the existence of the DNS, so he doesn't
have any complaints about the DNS zone files.

Yeah, like we'd name a host insecure-payroll-data.hackme.bugger.edu.
Or maybe unencrypted-credit-card-number-storage.bugger.edu.

Common sense isn't.
- Brian

[Roger Burton West]

Post by Brian Kantor
Yeah, like we'd name a host insecure-payroll-data.hackme.bugger.edu.
Or maybe unencrypted-credit-card-number-storage.bugger.edu.

When I joined $ISP, there were hosts called pr0n.$ISP.net and
warez.$ISP.net. We put them in the external DNS _specifically because_
they were on a network with no external access.

R

[Brian Kantor]

Post by Roger Burton West
When I joined $ISP, there were hosts called pr0n.$ISP.net and
warez.$ISP.net. We put them in the external DNS _specifically because_
they were on a network with no external access.

[karoshi.ucsd.edu] 2098 : host warez
warez.ucsd.edu is an alias for localhost.ucsd.edu.
localhost.ucsd.edu has address 127.0.0.1

[Paul Martin]

Post by Brian Kantor

Post by Roger Burton West
When I joined $ISP, there were hosts called pr0n.$ISP.net and
warez.$ISP.net. We put them in the external DNS _specifically because_
they were on a network with no external access.

[karoshi.ucsd.edu] 2098 : host warez
warez.ucsd.edu is an alias for localhost.ucsd.edu.
localhost.ucsd.edu has address 127.0.0.1

Great minds think alike, etc:

$ host warez.zetnet.co.uk
warez.zetnet.co.uk A 127.0.0.1

Actually, this year's April Fool status message was built around that.

[David P. Murphy]

Post by Paul Martin
$ host warez.zetnet.co.uk
warez.zetnet.co.uk A 127.0.0.1
Actually, this year's April Fool status message was built around that.

What, the one meant for the Scientologists?

ObPendant: since "-ology" refers to "the science of" . . .

ok
dpm

[Tanuki]

Post by David P. Murphy

Post by Paul Martin
$ host warez.zetnet.co.uk
warez.zetnet.co.uk A 127.0.0.1
Actually, this year's April Fool status message was built around that.

What, the one meant for the Scientologists?
ObPendant: since "-ology" refers to "the science of" . . .

{Maureen Lipman playing "Beattie"[1]}
"An Ology! He gets an ology and he says he's failed. You get an ology,
you're a scientist!"
{/Maureen Lipman}


[1]http://en.wikipedia.org/wiki/Beatrice_Bellman

[Garrett Wollman]

Post by David P. Murphy
ObPendant: since "-ology" refers to "the science of" . . .

Except that it doesn't.

-GAWollman

[Mike Andrews]

Post by Brian Kantor

Post by Menno
What is it about hostnames that so gets up people's snouts?
And then, with the naming convention well and truly established, the
DNS servers are duly buggered up, so that everybody uses the IP
addresses anyway.

And then the auditor comes in and insists that the hostnames must be
kept secret (i.e., not in the hosts table) because they might serve
to direct an attacker to a weak host, so you now have to encrypt
the hosts table. Of course, said auditor received his computer security
from a book published before the existence of the DNS, so he doesn't
have any complaints about the DNS zone files.
Yeah, like we'd name a host insecure-payroll-data.hackme.bugger.edu.
Or maybe unencrypted-credit-card-number-storage.bugger.edu.

OhGods, not there, too!

I'd call our auditors muttonheaded, except that it'd insult all sheep.

Remember, our auditors insisted on NETBIOS access to the mainframe.

Post by Brian Kantor
Common sense isn't.
- Brian

It isn't even _un_common; it's much rarer than that.

[Menno]

Post by Mike Andrews
OhGods, not there, too!
I'd call our auditors muttonheaded, except that it'd insult all sheep.
Remember, our auditors insisted on NETBIOS access to the mainframe.

Ahhhhh... Our good friends the security auditors, always ready to
point out the little details that might have escaped us while we were
trying to patch up the gaping big holes that our Auditing friends
completely failed to notice, like about half a million files to which
chmod 777 had been applied.

One of their remarks was that certain of our printer servers and tape
libraries were running FTP servers. And as we all know, FTP is
inherently insecure because ZzzZZz sends passwZZzzzzzZ plaintexZZZZzz
Zzzzzcket sniffZZZzz take over your whole print server. Dammit, I
actually had to *action* that at some point. Which I did with all due
procrastination until my IT Director, bless him, made it go away.

Feed them drunk and export them off the premises. Works a treat. Not
only on security auditors, but also on ISO-9000 scum.

[kyriacos Sakkas]

Post by Menno

Post by Mike Andrews
OhGods, not there, too!
I'd call our auditors muttonheaded, except that it'd insult all sheep.
Remember, our auditors insisted on NETBIOS access to the mainframe.

...

[stuff removed]

Post by Menno
...
Feed them drunk and export them off the premises. Works a treat. Not
only on security auditors, but also on ISO-9000 scum.

^
+------|-------+
|Kill kill kill|
+--------------+

Thank you, bye

--
what it says in the headers

[Niklas Karlsson]

Post by kyriacos Sakkas

Post by Menno
...
Feed them drunk and export them off the premises. Works a treat. Not
only on security auditors, but also on ISO-9000 scum.

^
+------|-------+
|Kill kill kill|
+--------------+

Well, yes.

This group seems to be in "remind Niklas of the sucky sides of his old
job so he appreciates his current one more" mode lately.

Niklas

[Garrett Wollman]

Post by Brian Kantor
And then the auditor comes in and insists that the hostnames must be
kept secret (i.e., not in the hosts table)

What is this "hosts table" of which you speak?

-GAWollman

[chris+ (Chris Suslowicz)]

Post by Menno
Engfeh.
Just spent these last two days renaming all of the hosts in my nice
test/development domain. InconCEIvable how many places that hostname
pops up in. They were originally named after cartoon characters, but
it was deemed that someone might see it and experience a brief moment
of glee, which is clearly inappropriate.
So now we named the buggers after Artists: Leonardo, Donatello,
Raphael, Michelangelo and the famous 14th century Dutch renaissance
painter Hans Splinter. And thus is my brief moment of glee preserved.

And the Impressionist: Hertz VanRental....

There was a PC upgrade project at Ork entitled "Raphael". My folder
of emailed bullethole^H^H^H^Hins and stuff was "Donatello", which
provoked the surprised comment from the person doing the upgrades
"But how did you _know_ it was named after the Ninja Turtles?"

"You're not cleared for that information, sorry."

(And I continued pounding away at the dumb terminal....)

Post by Menno
What is it about hostnames that so gets up people's snouts? I've seen
horrid constructions. There's the school of thought that thinks it
would be a good idea to stuff as much information as possible into the
hostname, leading to names like "rs6000-aix51-2002-03-01-1". (Which of
course runs AIX 5.3, because renaming it after the upgrade would have
been Difficult). Then, there is the "We need to know that it belongs
to Our Company" school of thought, leading to
MyCompany0001.mycompany.com. Or the "Lack of imagination" school of
thought: "box0001".

They're all a bunch of tossers who think the Political Officer will
find out and send them on sensitivity training courses or something.

Or just tossers, of course.

(Do not draw the 'fluffy cloud' as a sheep on your notwork diagram, either.)

Post by Menno
And then, with the naming convention well and truly established, the
DNS servers are duly buggered up, so that everybody uses the IP
addresses anyway.

Cue Tanuki .sig....

Post by Menno
Anyways, everything is working again. Time for beer and merriment.

Splendid idea, I shall add fireworks (tomorrow).

Chris.

[Satya]

Post by Menno
pops up in. They were originally named after cartoon characters, but
it was deemed that someone might see it and experience a brief moment
of glee, which is clearly inappropriate.

Someone thought displaying the date as a stardate on our intranet "portal"
page was "not the image we might want to present".

So now they don't get a date at all. They do get an IP address. I'm sure
someone will complain soon.

Post by Menno
So now we named the buggers after Artists: Leonardo, Donatello,
Raphael, Michelangelo and the famous 14th century Dutch renaissance
painter Hans Splinter. And thus is my brief moment of glee preserved.

I had to think about that one, then I imagined, what would be the
hostname of a box named after Hans Splinter....

Some people don't have a shred of imagination. In the evolution of
intelligence, these people must be mutants. They just want to stay in
their shell.

[Satya]

Post by Satya
Someone thought displaying the date as a stardate on our intranet "portal"
page was "not the image we might want to present".

This was two months after we started displaying it, by the way.

[Peter H. Coffin]

Post by Satya

Post by Satya
Someone thought displaying the date as a stardate on our intranet "portal"
page was "not the image we might want to present".

This was two months after we started displaying it, by the way.

It's an intranet. Usually that means "not seen by the general public".
How much is a stardate going to shatter the illusions of the people whom
actually have to work there? "Gee, I don't want this to be my employer;
they're so unprofessional"?

[jester]

On Sat, 24 Jun 2006 11:13:01 -0500, Peter H. Coffin

Post by Peter H. Coffin
"Gee, I don't want this to be my employer;
they're so unprofessional"?

Some people do think like that.

How hard is it to syncronise the clocks, make DNS work both ways, and set
up a new starter's necessary accounts *and* tell them about them?

[Satya]

Post by Peter H. Coffin

Post by Satya

Post by Satya
Someone thought displaying the date as a stardate on our intranet "portal"
page was "not the image we might want to present".

This was two months after we started displaying it, by the way.

It's an intranet. Usually that means "not seen by the general public".
How much is a stardate going to shatter the illusions of the people whom
actually have to work there? "Gee, I don't want this to be my employer;
they're so unprofessional"?

Maybe I shouldn't have said intranet.

a) It's a sniversity
b) It's a web page that anyone can go to. People with accounts can log
in and customize it a bit.
c) It's the default page for the million or so[0] workstations in the
[censored], which is where all the money comes from.
d) The stardate wasn't that obvious, despite being at the top of the
page.
e) We don't have MP3s of Kirk singing.
f) Complainer claims to be a Trek fan.
g) We have bigger things to be concerned about.
g) ii) Like 508 compliance. As in, "not compliant" "What's 508?"[1]



[0] That's what it sounds like, anyway. And feels like between 0800 and
0900. And when we inadvertently kill the system and the helpdesk goes
into a panic. Why the hell did we hardcode a cookie expiry of 14th June
2006 anyway?
[1] Notice the lack of mention of standards compliance.

[n***@buffy.sighup.org.uk]

Post by Satya
f) Complainer claims to be a Trek fan.

Words are free, it's deeds that count.

[Aquarion]

Post by Peter H. Coffin

Post by Satya

Post by Satya
Someone thought displaying the date as a stardate on our intranet "portal"
page was "not the image we might want to present".

This was two months after we started displaying it, by the way.

It's an intranet. Usually that means "not seen by the general public".
How much is a stardate going to shatter the illusions of the people whom
actually have to work there? "Gee, I don't want this to be my employer;
they're so unprofessional"?

So we had a meeting.

One... no, two steps backwards. We used to be a place where you had
people who needed stuff done, and people who did stuff. As the company
grew, this worked less, so they split the company into three, a group of
people who needed stuff done for clients, a group of people who did
stuff, and a nice thick wodge of greasy management between the two. A
work sandwich, if you will.

At the time, I was a Sysadmin/Developer. I am, in fact, the Other
Sysadmin. The man who is Lead Sysadmin was turned into a
Developer/Sysadmin. notice the reversal? It means I get to do all the
sysadmin, get behind on my development, but can't finish anything
because the man who has to sign it off is booked solid with development
work. This is Problem One.

Problem Two is more fun. It's where I am, as of a week today, not a
sysadmin anymore, because they've hired a full time sysadmin who isn't
also a developer. I'd like that job, but wasn't offered it.
Incidentally, time in lieu and being paid for being on call and that
kind of basic stuff? That came in as of friday. That is Problem Two.

Problem Three is that I have a reduntant array of expensive sell dervers
sitting across four or five desks (We have no racking) needing to be set
up so we can ship them down to $DATACENTRE in the Smoke. (Which I can't
finish because I need to get sign-off on the differences between them
and our usual dev boxes. See above).

So we had a meeting.

At this meeting I was told that I would not be allowed to name these
machines. Our current naming scheme (which is, in effect, that we don't
have one. Machines built at the same time for the same purpose follow a
small scheme, all others are random), because the people who don't need
to know which machine does what can't tell which machine does what when
the machine stops doing it and sulks instead. All machines will now be
called "${function}{$x}" where x is an int, and function isn't.
webhosts1-3 are almost ready, but I couldn't tell you which will do
what, and apppilotserver3 has already been redesignated to do something
instead of hosting pilots (not that kind).

Problem Four is that my unemployment insurance doesn't cover resigning.

[Menno]

Post by Aquarion
All machines will now be
called "${function}{$x}" where x is an int, and function isn't.
webhosts1-3 are almost ready, but I couldn't tell you which will do
what, and apppilotserver3 has already been redesignated to do something
instead of hosting pilots (not that kind).

I do not approve of attaching meaning to hostnames. That's what CNAMEs
are there for. Hostnames should be utterly bereft of any hint of what
might be going on on the hosts in question, so that when they start
doing things different from their original job description, you don't
need to dig through everybody's host files, SSH keys and whatnot, to
change the hostname, or (as is more common) end up with a name scheme
vaguely reminiscent of the names of the Crude Artisans in "Lords and
Ladies".

And now for my next trick, I shall go and find a choir and preach to
it.

[Richard Bos]

Post by Menno

Post by Aquarion
All machines will now be
called "${function}{$x}" where x is an int, and function isn't.
webhosts1-3 are almost ready, but I couldn't tell you which will do
what, and apppilotserver3 has already been redesignated to do something
instead of hosting pilots (not that kind).

I do not approve of attaching meaning to hostnames. That's what CNAMEs
are there for.

Butbut... CNAMEs... you can't be suggesting that we use *gasp* a name
server? Surely you wouldn't deprive us of the amusement of 2000-line[1]
/etc/hosts, on each of the 2000 machines, to be updated each time any
machine changes function, hardware, location, or - oh joy - IP address?
And oh, the fun you have when you forget to update a machine!

Richard

[1] never mind how many individual names

[Maarten Wiltink]

"Richard Bos" <***@xs4all.nl> wrote in message news:***@news.xs4all.nl...
[...]

Post by Richard Bos
Butbut... CNAMEs... you can't be suggesting that we use *gasp* a name
server? Surely you wouldn't deprive us of the amusement of 2000-line[1]
/etc/hosts, ...

As a matter of historical interest, does anybody here who was around to
witness DNS being introduced, have an idea of the distribution of hosts
file lengths among the then-extant population?

Tebrgwrf,
Maarten Wiltink

[Mike Andrews]

Post by Maarten Wiltink
[...]

Post by Richard Bos
Butbut... CNAMEs... you can't be suggesting that we use *gasp* a name
server? Surely you wouldn't deprive us of the amusement of 2000-line[1]
/etc/hosts, ...

As a matter of historical interest, does anybody here who was around to
witness DNS being introduced, have an idea of the distribution of hosts
file lengths among the then-extant population?

I had a hosts file about 150 KBytes long, back around 1986, given me
by a friend at ou.edu who preferred it to waiting for DNS packets to
make their way through ou's terribly congested network.

[Randy the Random]

Post by Mike Andrews
I had a hosts file about 150 KBytes long, back around 1986, given me
by a friend at ou.edu who preferred it to waiting for DNS packets to
make their way through ou's terribly congested network.

My personal hosts file was no more than 300 lines. That was before I
made it to the university and the world changed.

[Richard Gadsden]

Post by Menno
I do not approve of attaching meaning to hostnames. That's what CNAMEs
are there for.

There are two possible conclusions to this:

Either hostnames are arbitrary, name your host whatever you like.

Or hostnames must be meaningless, so we're going to number all the
hosts. Oh joy, I now have to remember what srv84 does.

--

[TimC]

On 2006-06-26, Richard Gadsden (aka Bruce)

Post by Richard Gadsden

Post by Menno
I do not approve of attaching meaning to hostnames. That's what CNAMEs
are there for.

Either hostnames are arbitrary, name your host whatever you like.
Or hostnames must be meaningless, so we're going to number all the
hosts. Oh joy, I now have to remember what srv84 does.

Why would you want to do that? What's wrong with 210.87.42.227?

[Stuart Lamble]

Post by TimC
On 2006-06-26, Richard Gadsden (aka Bruce)

Post by Richard Gadsden
Oh joy, I now have to remember what srv84 does.

Why would you want to do that? What's wrong with 210.87.42.227?

Absolutely nothing, except that it's 210.42.87.227 that you want.

[Peter Corlett]

Aquarion <***@suespammers.org> wrote:
[...]

Post by Aquarion
Problem Four is that my unemployment insurance doesn't cover resigning.

Does it cover constructive dismissal?

[Aquarion]

Post by Peter Corlett
[...]

Post by Aquarion
Problem Four is that my unemployment insurance doesn't cover resigning.

Does it cover constructive dismissal?

It depends if I can afford to argue about it. I believe the wording of
my contract states that I am a sysadmin in association with others. My
job title doesn't change, it's just my rank in the "Who we gonna call?"
list goes from Ghostbusters to animal control.

[n***@buffy.sighup.org.uk]

Post by Menno
What is it about hostnames that so gets up people's snouts? I've seen
horrid constructions. There's the school of thought that thinks it
would be a good idea to stuff as much information as possible into the
hostname, leading to names like "rs6000-aix51-2002-03-01-1". (Which of
course runs AIX 5.3, because renaming it after the upgrade would have
been Difficult). Then, there is the "We need to know that it belongs
to Our Company" school of thought, leading to
MyCompany0001.mycompany.com. Or the "Lack of imagination" school of
thought: "box0001".

Oh yes, the "lets-encode-data-in-the-name-because-databases-are-difficult"
school of naming. We have some servers with carefully constructed names so
you should be able to tell which servers are production and which are test
and where each is located. Murphy ensures that test servers are running
production services and the hardware has been moved (or not moved) so it is
no longer in the correct location.

Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

[chris+ (Chris Suslowicz)]

Post by n***@buffy.sighup.org.uk
Oh yes, the "lets-encode-data-in-the-name-because-databases-are-difficult"
school of naming. We have some servers with carefully constructed names so
you should be able to tell which servers are production and which are test
and where each is located. Murphy ensures that test servers are running
production services and the hardware has been moved (or not moved) so it is
no longer in the correct location.
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Wallace, Gromit & Sean. The operations manager *did* have a sense of humour,
and the servers concerned (AS/400s, I think) each had a hand-knitted mascot
(produced by one of the support staff) mascot on top.

Chris.

[Paul Martin]

Post by chris+ (Chris Suslowicz)
Wallace, Gromit & Sean. The operations manager *did* have a sense of humour,
and the servers concerned (AS/400s, I think) each had a hand-knitted mascot
(produced by one of the support staff) mascot on top.

What's a mascot mascot?

[chris+ (Chris Suslowicz)]

Post by Paul Martin

Post by chris+ (Chris Suslowicz)
Wallace, Gromit & Sean. The operations manager *did* have a sense of humour,
and the servers concerned (AS/400s, I think) each had a hand-knitted mascot
(produced by one of the support staff) mascot on top.

What's a mascot mascot?

One of those irritating repeated word typo's that is very difficult to spot.

HTH

(Thinking about it, they may have been Gromit, Feathers and Sean, as the
Wallace, Wendolyn & Preston knitting patterns came out later. The customer
has since converted to all midrange and taken the work back in-house,
after having a ceremonial switch-off of the dinosaur and a "thanks for
all your work over the years" event. They were back a week later asking
if we could fire the beast back up and run a final 'comparison report'
that someone had forgotten to do.... by that time the dinosaur had been
disconnected, but the final wipe of their raid box hadn't been started
and they were even more pleased to get the report the following day -
due to the wonders of RFPBA qverpgbef and a spare dinosaur.[1])

Chris.
[1] Unlike another customer who transferred to a (cheaper) outsourcing
TLA company, who called back frantically 10 days later asking if
it was possible to take the work back again "if we broke our contract
with the new supplier"[2]. Sadly, it wasn't possible - all the kit
had been sanitized, disconnected, and returned to the suppliers.
[2] There was a reason they could undercut us: *everything* seemed to
have become an "extra cost item", and they slashed the support staff
to the bone in order to make their idea of a profit. Furrfu!

[Zebee Johnstone]

In alt.sysadmin.recovery on Sun, 25 Jun 2006 12:37:27 +0100

Post by chris+ (Chris Suslowicz)
[2] There was a reason they could undercut us: *everything* seemed to
have become an "extra cost item", and they slashed the support staff
to the bone in order to make their idea of a profit. Furrfu!

Isn't that how all outsourcing works?

It took the previously oursourced and now insourced admins from HP
that we hired up to 2 weeks to lose the "I can't do that, it's not in
scope" mindset. I heard a number of "at last I can fix this cockup!"
comments too.

Zebee

[chris+ (Chris Suslowicz)]

Post by Zebee Johnstone
In alt.sysadmin.recovery on Sun, 25 Jun 2006 12:37:27 +0100

Post by chris+ (Chris Suslowicz)
[2] There was a reason they could undercut us: *everything* seemed to
have become an "extra cost item", and they slashed the support staff
to the bone in order to make their idea of a profit. Furrfu!

Isn't that how all outsourcing works?

Our variety doesn't seem to (it's more of an 'economy of scale' thing).

Post by Zebee Johnstone
It took the previously oursourced and now insourced admins from HP
that we hired up to 2 weeks to lose the "I can't do that, it's not in
scope" mindset. I heard a number of "at last I can fix this cockup!"
comments too.

Ah yes, Pewlett Hackard, and their utterly fscking dreadful problem
ticketing system that I will be fighting all next week. (Not to mention
the other four incompatible and bizarre/baroque/b0Rk3N ones: ARSe,
PN-HFQ, Inagvir and Gvibyv Freivpr Qrfx (The latter is quite nice,
having been developed to the customer requirements, but is now
substantially read-only having been forcibly supplanted by Pewlett
Hackard's choice: the infamous Pynevsl from Abegry Abgjbexf.)

Argh! Another 11 hours 51 minutes before I have to deal with that crap,
where's the Balvenie?

Chris.

[Zebee Johnstone]

In alt.sysadmin.recovery on Sun, 25 Jun 2006 21:40:09 +0100

Post by chris+ (Chris Suslowicz)
substantially read-only having been forcibly supplanted by Pewlett
Hackard's choice: the infamous Pynevsl from Abegry Abgjbexf.)

even Rotted it makes my head hurt.

"what totally irrelevant cause and solution codes should I use this
time?" - which is a random choice of things I hate about Pynevsl.

It's Abegry I hve to blame eh? Shouldn't surprise me, given the VPN
we have to use, and the software for same.

Why is their client software user interface so massively different on
OSX, Windows, and Linux? Why is the OSX one so bad? And what
horrors is it perpetrating so that it only works in IE or Mozilla?

Zebee

[chris+ (Chris Suslowicz)]

Post by Zebee Johnstone
In alt.sysadmin.recovery on Sun, 25 Jun 2006 21:40:09 +0100

Post by chris+ (Chris Suslowicz)
substantially read-only having been forcibly supplanted by Pewlett
Hackard's choice: the infamous Pynevsl from Abegry Abgjbexf.)

even Rotted it makes my head hurt.
"what totally irrelevant cause and solution codes should I use this
time?" - which is a random choice of things I hate about Pynevsl.
It's Abegry I hve to blame eh?

Certainly the bukkake screen on startup proclaims they are responsible.
I think Pewlett Hackard are irresponsible for the total lack of
customisation of this one: it's *still* a bloody Telco service desk tool!

(It also has an 'interesting' timeout failure mode, where it cheerily
tells you the database has gone away and to "Exit Immediately". Of
course, the only way to do that on a 'doze box is via the task mangler....

Post by Zebee Johnstone
Shouldn't surprise me, given the VPN
we have to use, and the software for same.
Why is their client software user interface so massively different on
OSX, Windows, and Linux? Why is the OSX one so bad? And what
horrors is it perpetrating so that it only works in IE or Mozilla?

ASS, of course.

<reaches for the Balvenie again>

Chris.

[Niklas Karlsson]

Post by Zebee Johnstone
In alt.sysadmin.recovery on Sun, 25 Jun 2006 21:40:09 +0100

Post by chris+ (Chris Suslowicz)
substantially read-only having been forcibly supplanted by Pewlett
Hackard's choice: the infamous Pynevsl from Abegry Abgjbexf.)

even Rotted it makes my head hurt.
"what totally irrelevant cause and solution codes should I use this
time?" - which is a random choice of things I hate about Pynevsl.
It's Abegry I hve to blame eh? Shouldn't surprise me, given the VPN
we have to use, and the software for same.

Graaagh, the memories. For most of the time I spent Bobbing for
WeSellBridges, we used Pynevsl to track support cases. The time came
when a new solution was to be implemented, and I figured it couldn't
possibly be worse than what we already had.

Needless to say, sure enough, it was worse and got just about everything
wrong that Pynevsl had actually got _right_. And to top it all off it
was Wnin-based.

Among its more egregious failings from my team's perspective was that
you could only view one case at a time, which is of course great when
you're handling several of them at a time...

We got around this by having multiple sessions open at a time, until we
were told in no uncertain terms Not To Do That, because apparently it
screwed with the backend in unspecified ways.

Yes, ASS, but not all software manages to be such festering piles of
rotten llama dung as most ticketing systems appear to be. That's before
I even get started on the decision to actually _buy_ the new system with
knowledge of its limitations. (Well, they _claimed_ to have known.
That's not saying much, of course.) I'd have told them to put down their
crack pipes and open the window, but venting the fumes of that many
people at once would probably have violated some local environmental
ordinance.

Every time I have one of those "why did I sign up for this job?" days at
my new source of unrecovery, I think to myself "at least I don't have to
put up with WeSellBridges' infernal maggot-ridden excuse for a new
ticketing system anymore."

The owner of my local pub may thank you two for reminding me of it.

Niklas

[Anthony de Boer - USEnet]

[ ticketing systems that suck ]
Among its more egregious failings from my team's perspective was that
you could only view one case at a time, which is of course great when
you're handling several of them at a time...
We got around this by having multiple sessions open at a time, until we
were told in no uncertain terms Not To Do That, because apparently it
screwed with the backend in unspecified ways.

Oh, *JOY*. That's probably the point at which you stick with the one
ticket you have open until you can close it, before you dive into the
queue for the next.

Should be educational, at least in the sense of seeing what workaround
TPTB come up with to cover this misfeature and keep things rolling.

[Mike Andrews]

Post by Anthony de Boer - USEnet

[ ticketing systems that suck ]
Among its more egregious failings from my team's perspective was that
you could only view one case at a time, which is of course great when
you're handling several of them at a time...
We got around this by having multiple sessions open at a time, until we
were told in no uncertain terms Not To Do That, because apparently it
screwed with the backend in unspecified ways.

Oh, *JOY*. That's probably the point at which you stick with the one
ticket you have open until you can close it, before you dive into the
queue for the next.
Should be educational, at least in the sense of seeing what workaround
TPTB come up with to cover this misfeature and keep things rolling.

If so, you'll come out a changed man.

[Mike Andrews]

Anthony de Boer - USEnet <***@leftmind.net> sigged:

< "Is someone piping me through sed without my knowledge?" -- Malcolm Ray >

If so, you'll come out a changed man.

[Cameron Biggart]

Post by Niklas Karlsson
I'd have told them to put down their
crack pipes and open the window, but venting the fumes of that many
people at once would probably have violated some local environmental
ordinance.

Not to mention the fuel cost of getting the first one alight enough to have
sustained combustion On the plus side you could probably use them as the
ignition source for the others on the pile. Getting rid of the ashes I
leave as an exercise for the reader.

[Roger Burton West]

Post by Zebee Johnstone
In alt.sysadmin.recovery on Sun, 25 Jun 2006 21:40:09 +0100

Post by chris+ (Chris Suslowicz)
substantially read-only having been forcibly supplanted by Pewlett
Hackard's choice: the infamous Pynevsl from Abegry Abgjbexf.)

even Rotted it makes my head hurt.

I'm just surprised Erzrql wasn't in the list. Has it been renamed?

Post by Zebee Johnstone
Why is their client software user interface so massively different on
OSX, Windows, and Linux?

My bet: they had it written for Windows because that's the only OS
anyone uses - they asked _all_ their friends at the golf club. Then some
of their customers started whining, so they told the new guy "write a
Linux interface"...

R

[Mike Andrews]

Post by Roger Burton West

Post by Zebee Johnstone
In alt.sysadmin.recovery on Sun, 25 Jun 2006 21:40:09 +0100

Post by chris+ (Chris Suslowicz)
substantially read-only having been forcibly supplanted by Pewlett
Hackard's choice: the infamous Pynevsl from Abegry Abgjbexf.)

even Rotted it makes my head hurt.

I'm just surprised Erzrql wasn't in the list. Has it been renamed?

Erzrql is not good enough to be in the "bad" list; it's one of the
standards against which everything has to look better.

It appears to be named by contraries, like the anti-junkfax and
anti-spam bills, which really facilitate junk faxing and spamming.
Erzrql really only "remedies" sanity.

Corollary: If it's worse than Erzrql, then it doesn't exist. At least
I _HOPE_ it doesn't exist, and that I'm never proven wrong on that.

[Richard Bos]

Post by Mike Andrews
Erzrql is not good enough to be in the "bad" list; it's one of the
standards against which everything has to look better.
It appears to be named by contraries, like the anti-junkfax and
anti-spam bills, which really facilitate junk faxing and spamming.
Erzrql really only "remedies" sanity.
Corollary: If it's worse than Erzrql, then it doesn't exist. At least
I _HOPE_ it doesn't exist, and that I'm never proven wrong on that.

Conclusion: after expressing that hope, you will be proven wrong, and
sooner than you think.

Richard

[Graham Reed]

Post by Roger Burton West
I'm just surprised Erzrql wasn't in the list. Has it been renamed?

You're confusing the Company with the Product. The Product is ARS,
and was, indeed, in the list. (Though spelled as by one who has used
it: "ARSe".)

Theoretically, ARSe is merely a GUI drawing thingie that sits on top
of a user-defined schema. It would be possible, therefore, to produce
something that at least sucks a bit less.

But I've never heard of such a schema; they always suck ARSe.

[n***@buffy.sighup.org.uk]

Post by Graham Reed
Theoretically, ARSe is merely a GUI drawing thingie that sits on top
of a user-defined schema. It would be possible, therefore, to produce
something that at least sucks a bit less.

They've had four goes at a web interface and each has sucked. It seems that
there is a law of nature that says that ARS sucks at all levels for all time.

They haven't tried abusing AJAX or GWT yet so we can look forward to at
least two more releases that suck wombats through bucky tubes.

[Maarten Wiltink]

"Graham Reed" <***@pobox.com> wrote in message news:***@greed.dyndns.org...
[...]

Post by Graham Reed
Theoretically, ARSe is merely a GUI drawing thingie that sits on top
of a user-defined schema. It would be possible, therefore, to produce
something that at least sucks a bit less.

Hey, that sounds just like the thing I'm concocting at work right now.
People keep asking for new tricks that the software should be able to
turn^Wperform with the schema.

Post by Graham Reed
But I've never heard of such a schema; they always suck ARSe.

Oh, it's the _schema_ you find fault with for sucking. The schema isn't
strictly software. It's more like configuration, that is, data. Another
part that can suck. While the software never stops.

Tebrgwrf,
Maarten Wiltink

[Red Drag Diva]

On Mon, 26 Jun 2006 12:12:43 +0000 (UTC),
Roger Burton West <***@nospam.firedrake.org> wrote:

: I'm just surprised Erzrql wasn't in the list. Has it been renamed?


Yeah, "qvfrnfr".

[Richard Bos]

Post by chris+ (Chris Suslowicz)
Argh! Another 11 hours 51 minutes before I have to deal with that crap,
where's the Balvenie?

In my cupboard. And you can't have it - I want to have some left to
celebrate my birthday next week.

Richard

[chris+ (Chris Suslowicz)]

Post by Richard Bos

Post by chris+ (Chris Suslowicz)
Argh! Another 11 hours 51 minutes before I have to deal with that crap,
where's the Balvenie?

In my cupboard. And you can't have it - I want to have some left to
celebrate my birthday next week.

I have my own bottle (only the 10yo, but very nice), and spent yesterday
evening at Medium Brother's house: his lovely lady's birthday party. At
that point I actually _read_the_instructions_ on a 75mm (firework) shell:

"Light fuse before put in tube"

Jesus H. Christ! It had an instantaneous fuse, and I sincerely hope
nobody ever tried to follow those instructions[1].

Furrfu!

Chris.
[1] scorched fingers at the very least, then you have 2 seconds to
run away before the shell lying at your feet goes off. I must
see of the same 'instructions' are on their mines - they have
zero delay between lift and burst, so would be even less fun.

[Richard Bos]

Post by chris+ (Chris Suslowicz)

Post by Richard Bos

Post by chris+ (Chris Suslowicz)
Argh! Another 11 hours 51 minutes before I have to deal with that crap,
where's the Balvenie?

In my cupboard. And you can't have it - I want to have some left to
celebrate my birthday next week.

I have my own bottle (only the 10yo, but very nice),

Nice indeed. Mine has "Founder's Reserve" on it; I don't know whether
that means it's different from the normal 10yo, and frankly, after a
glass or so, I don't care any more. I do like their bottles, though.

Richard

[chris+ (Chris Suslowicz)]

Post by Richard Bos

Post by chris+ (Chris Suslowicz)

Post by Richard Bos

Post by chris+ (Chris Suslowicz)
where's the Balvenie?

In my cupboard. And you can't have it - I want to have some left to
celebrate my birthday next week.

I have my own bottle (only the 10yo, but very nice),

Nice indeed. Mine has "Founder's Reserve" on it; I don't know whether
that means it's different from the normal 10yo, and frankly, after a
glass or so, I don't care any more. I do like their bottles, though.

That's the one I've got. Slightly cheaper than the 12yo Doublewood,
and very nice indeed... I must do a comparative test sometime (but
not tonight as I'm on call).

Chris.

[Peter Corlett]

Chris Suslowicz <chris+***@suslowicz.org> wrote:
[...]

That's the one I've got. Slightly cheaper than the 12yo Doublewood, and
very nice indeed... I must do a comparative test sometime (but not tonight
as I'm on call).

I was somewhat pleased when I went into the Vintage House, ordered a bottle
of plain Balvenie, and I got a DoubleWood for the same low price.

At some point, I will find and unpack my glasses and not be a slob sipping
it straight out of the bottle. (Sorry, but I couldn't wait.)

[Jim]

Post by Peter Corlett
[...]

That's the one I've got. Slightly cheaper than the 12yo Doublewood, and
very nice indeed... I must do a comparative test sometime (but not tonight
as I'm on call).

I was somewhat pleased when I went into the Vintage House, ordered a bottle
of plain Balvenie, and I got a DoubleWood for the same low price.
At some point, I will find and unpack my glasses and not be a slob sipping
it straight out of the bottle. (Sorry, but I couldn't wait.)

I've done that as well (with an Irish whiskey, the name of which escapes me
at the moment). My excuse was: I was in a car (not driving) and I was in
non-trivial amounts of pain.

Sometimes needs must.

Jim

[Anthony de Boer - USEnet]

Post by chris+ (Chris Suslowicz)
...
the other four incompatible and bizarre/baroque/b0Rk3N ones: ARSe,

^^^^

Post by chris+ (Chris Suslowicz)
PN-HFQ, Inagvir and Gvibyv Freivpr Qrfx ...

And a fine stealth ROTism that is.

[Tanuki]

Post by Zebee Johnstone
In alt.sysadmin.recovery on Sun, 25 Jun 2006 12:37:27 +0100

Post by chris+ (Chris Suslowicz)
[2] There was a reason they could undercut us: *everything* seemed to
have become an "extra cost item", and they slashed the support staff
to the bone in order to make their idea of a profit. Furrfu!

Isn't that how all outsourcing works?
It took the previously oursourced and now insourced admins from HP
that we hired up to 2 weeks to lose the "I can't do that, it's not in
scope" mindset.

If you are ever faced with the prospect of negotiating contract-
amendments with HP Legal, I can suggest two possible courses of
action.

1]Load up on perception-altering drugs well in advance. HP's legal-
people are already living in a different world so you might as well
aim to meet them half-way.
2]Go into the meeting carrying belt-fed weapons, plenty of ammo, and
having practiced the phrase "my client is not prepared to negotiate
on this point" in front of the mirror.

[Alan J Rosenthal]

Post by Paul Martin

Post by chris+ (Chris Suslowicz)
and the servers concerned (AS/400s, I think) each had a hand-knitted mascot
(produced by one of the support staff) mascot on top.

What's a mascot mascot?

Obviously, something less valuable than a hand-knitted mascot mascot.

[Joe Zeff]

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

How about Ben and Jerry, or Pat and Mike? Lohman and Barkley would
also work, as would Hudson and Landry.

[Rob Adams]

On Fri, 23 Jun 2006 21:58:27 GMT, Joe Zeff

Post by Joe Zeff

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

How about Ben and Jerry, or Pat and Mike? Lohman and Barkley would
also work, as would Hudson and Landry.

Can you imagine the crap that Pat and Mike would turn out?

Rob.

[Richard Bos]

Post by n***@buffy.sighup.org.uk
Oh yes, the "lets-encode-data-in-the-name-because-databases-are-difficult"
school of naming. We have some servers with carefully constructed names so
you should be able to tell which servers are production and which are test
and where each is located. Murphy ensures that test servers are running
production services and the hardware has been moved (or not moved) so it is
no longer in the correct location.
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Since fail-over machines are traditionally fall-over machines, I suggest
Stanley and Oliver.

Richard

[Steve VanDevender]

Post by Richard Bos

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Since fail-over machines are traditionally fall-over machines, I suggest
Stanley and Oliver.

If they're particularly aggressive about their takeover behavior (such
as including a STONITH[1] device), I also suggest Itchy and Scratchy.

I once had an experience with a failover system where failover appeared
to work in tests, but in practice when it occurred the node that took
over on the failure of the primary then would then gently shut down all
of _its_ services about ten minutes after the failover, with the logs
suggesting that it believed it couldn't find itself so why was it
providing services? I dubbed this "existential crisis mode". I wonder
if I can get those two machines renamed to Sartre and Camus? Although
we found a way to prevent "existential crisis mode" after that.

[1] Shoot The Other Node In The Head.

[Shmuel (Seymour J.) Metz]

Post by Richard Bos
Since fail-over machines are traditionally fall-over machines, I
suggest Stanley and Oliver.

Larry, Curly and Moe. Perhaps also Shemp.

[stevo]

Post by Richard Bos

Post by n***@buffy.sighup.org.uk
Oh yes, the "lets-encode-data-in-the-name-because-databases-are-difficult"
school of naming. We have some servers with carefully constructed names so
you should be able to tell which servers are production and which are test
and where each is located. Murphy ensures that test servers are running
production services and the hardware has been moved (or not moved) so it is
no longer in the correct location.
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Since fail-over machines are traditionally fall-over machines, I suggest
Stanley and Oliver.

For some reason I parsed fall-over as stand-over and immediately thought
of Ronnie and Reggie.

[Stuart Lamble]

Post by stevo

Post by Richard Bos

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Since fail-over machines are traditionally fall-over machines, I suggest
Stanley and Oliver.

For some reason I parsed fall-over as stand-over and immediately thought
of Ronnie and Reggie.

Which immediately suggests to me naming systems "barker" and "corbett".

[Mike Andrews]

Post by Stuart Lamble

Post by stevo

Post by Richard Bos

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Since fail-over machines are traditionally fall-over machines, I suggest
Stanley and Oliver.

For some reason I parsed fall-over as stand-over and immediately thought
of Ronnie and Reggie.

Which immediately suggests to me naming systems "barker" and "corbett".

I'm amazed that nobody has suggested "Burke" & "Hare" yet, and more so
after someone's (AdB's?) observation about machines that are set up to
shoot their partners in the head as part of the process of taking up
the load.

ObLastWeekend: ARRL Field day. I ran the 6 meter station, craftily
letting the members' kids do the actual operations
while I just acted as control operator. We got 109
QSOs on 6, mostly at 50.170, where we squatted and had
as many as 5 stations responding to CQs. The 20m phone
station was mostly operated by YL club members, and had
700 QSOs when they stopped at 1758Z yesterday. I am led
to observe that folks out in the world vastly prefer to
answer calls from ladies and kids. I also observe that
having the kids operate and log is a lot of fun for us
and for the kids, and sure seems to have _these_ 6 kids
hooked. Youngest was 5 YO, the oldest 13.

[Anthony de Boer - USEnet]

Post by Mike Andrews
I'm amazed that nobody has suggested "Burke" & "Hare" yet, and more so
after someone's (AdB's?) observation about machines that are set up to
shoot their partners in the head as part of the process of taking up
the load.

I may have previously reported the technique here, after attending an OLS
talk by Alan Robertson in which he explained the term STONITH, complete
with attendant dramatics.

http://linux-ha.org/STONITH pops right up with a trivial search.

``Normally, when an HA system declares a node as dead, it is merely
speculating that it is dead. STONITH takes that speculation and makes
it reality.''

[Steve VanDevender]

Post by Anthony de Boer - USEnet

Post by Mike Andrews
I'm amazed that nobody has suggested "Burke" & "Hare" yet, and more so
after someone's (AdB's?) observation about machines that are set up to
shoot their partners in the head as part of the process of taking up
the load.

I may have previously reported the technique here, after attending an OLS
talk by Alan Robertson in which he explained the term STONITH, complete
with attendant dramatics.

I think it was my mention of it a few days ago that might have inspired
Mike.

Although after looking up the story of Burke and Hare, the version I
read did not end with one of them shooting the other in the head.

With all the duels fought through history, I think at least one would
have ended with both duelists simultaneously shooting each other in the
head, and the names of those duelists would be extremely appropriate for
a failover pair equipped with STONITH devices.

Post by Anthony de Boer - USEnet
http://linux-ha.org/STONITH pops right up with a trivial search.
``Normally, when an HA system declares a node as dead, it is merely
speculating that it is dead. STONITH takes that speculation and makes
it reality.''

If only the rest of their documentation were so pithy and cogent. I
have to maintain some of these and despite the quantity of their
documentation I have still had to resort to experiment to figure out
what many things _actually_ do.

[Zebee Johnstone]

In alt.sysadmin.recovery on 26 Jun 2006 22:13:24 -0700

Post by Steve VanDevender
With all the duels fought through history, I think at least one would
have ended with both duelists simultaneously shooting each other in the
head, and the names of those duelists would be extremely appropriate for
a failover pair equipped with STONITH devices.

ON the other hand, you could name them Bruce and Dorset after the
duellists mentioned in Frank Lurz's "The Dubious Quick Kill"
http://www.classicalfencing.com/articles/bloody.php

I've met cluster servers that get really confused about which one is
really dead....

"Take for example the case of the duel fought in 1613 between the Earl
of Dorset and Lord Edward Bruce.

According to the Earl's account, he received a rapier-thrust in the
right nipple which passed "level through my body, and almost to my back."
Seemingly unaffected, the Earl remained engaged in the combat for some
time. The duel continued with Dorset going on to lose a finger while
attempting to disarm his adversary manually. Locked in close quarters,
the two struggling combatants ultimately ran out of breath. According
to Dorset's account, they paused briefly to recover, and while catching
their wind, considered proposals to release each other's blades. Failing
to reach an agreement on exactly how this might be done, the seriously
wounded Dorset finally managed to free his blade from his opponent's grasp
and ultimately ran Lord Bruce through with two separate thrusts. Although
Dorset had received what appears to have been a grievous wound that,
in those days, ought to have been mortal, he not only remained active
long enough to dispatch his adversary, but without the aid of antibiotics
and emergency surgery, also managed to live another thirty-nine
years."

Zebee

[anon]

Post by stevo

Post by Richard Bos

Post by n***@buffy.sighup.org.uk
Oh yes, the "lets-encode-data-in-the-name-because-databases-are-difficult"
school of naming. We have some servers with carefully constructed names so
you should be able to tell which servers are production and which are test
and where each is located. Murphy ensures that test servers are running
production services and the hardware has been moved (or not moved) so it is
no longer in the correct location.
Then they look at me funny when I suggest that two machines connected in a
fail-over cluster should be called Tom and Jerry or Bill and Ben. I have a
tiny ambition to get two systems called Don and Phil.

Since fail-over machines are traditionally fall-over machines, I suggest
Stanley and Oliver.

For some reason I parsed fall-over as stand-over and immediately thought
of Ronnie and Reggie.

That's only going to work if they are Crays.

[Maarten Wiltink]

<***@buffy.sighup.org.uk> wrote in message news:449c59a3$0$661$***@news.aaisp.net.uk...
[...]

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected
in a fail-over cluster should be called Tom and Jerry or Bill and Ben.
I have a tiny ambition to get two systems called Don and Phil.

Don and Phil seems a particularly fruitful pair. Two singers, two
commercial priests, and two Blackbird jockeys (those shoes!).

Tebrgwrf,
Maarten Wiltink

[Mike Andrews]

Post by Maarten Wiltink
[...]

Post by n***@buffy.sighup.org.uk
Then they look at me funny when I suggest that two machines connected
in a fail-over cluster should be called Tom and Jerry or Bill and Ben.
I have a tiny ambition to get two systems called Don and Phil.

Don and Phil seems a particularly fruitful pair. Two singers, two
commercial priests, and two Blackbird jockeys (those shoes!).

So what sorts of shoes do SR-71A drivers wear, anyway?

[Maarten Wiltink]

Post by Mike Andrews

Post by Maarten Wiltink

[...] I have a tiny ambition to get two systems called Don and Phil.

Don and Phil seems a particularly fruitful pair. Two singers, two
commercial priests, and two Blackbird jockeys (those shoes!).

So what sorts of shoes do SR-71A drivers wear, anyway?

Can you guess how I found Don & Phil, Don & Phil, and Don & Phil?

Tebrgwrf,
Maarten Wiltink

[alt]

Post by n***@buffy.sighup.org.uk

Post by Menno
What is it about hostnames that so gets up people's snouts? I've seen
horrid constructions. There's the school of thought that thinks it
would be a good idea to stuff as much information as possible into
the hostname, leading to names like "rs6000-aix51-2002-03-01-1".
(Which of course runs AIX 5.3, because renaming it after the upgrade
would have been Difficult). Then, there is the "We need to know that
it belongs to Our Company" school of thought, leading to
MyCompany0001.mycompany.com. Or the "Lack of imagination" school of
thought: "box0001".

Oh yes, the
"lets-encode-data-in-the-name-because-databases-are-difficult" school
of naming. We have some servers with carefully constructed names so
you should be able to tell which servers are production and which are
test and where each is located. Murphy ensures that test servers are
running production services and the hardware has been moved (or not
moved) so it is no longer in the correct location.
Then they look at me funny when I suggest that two machines connected
in a fail-over cluster should be called Tom and Jerry or Bill and Ben.
I have a tiny ambition to get two systems called Don and Phil.

Not a fail-over cluster, but two hostnames for different IPs on the same
box:

pain
suffering

[Roger Burton West]

Post by alt
Not a fail-over cluster, but two hostnames for different IPs on the same
pain
suffering

And here I thought it would be:

Loss% Snt Last Avg Best Wrst StDev
1. fear 0.0% 20 0.4 0.4 0.4 0.9 0.1
2. anger 0.0% 20 0.6 0.6 0.6 0.8 0.0
3. hate 0.0% 20 18.1 42.9 15.2 121.0 30.4
4. suffering 15.0% 20 107.1 42.8 15.2 113.7 32.9
5. ??? 100.0 20 0.0 0.0 0.0 0.0 0.0

R

[Peter Corlett]

alt <***@lazyeyez.net> wrote:
[...]

Post by alt
Not a fail-over cluster, but two hostnames for different IPs on the same
box: pain suffering

I used "balti" and "tandoori" for a similar purpose, after noticing that
balti houses and tandoori places tend to be next door to each other and
probably under the same management.

[Brian Kantor]

Post by alt
Not a fail-over cluster, but two hostnames for different IPs on the same
box: pain suffering

fear & loathing

with apologies to Uncle Duke.
- Brian

[chris+ (Chris Suslowicz)]

Post by Brian Kantor

Post by alt
Not a fail-over cluster, but two hostnames for different IPs on the same
box: pain suffering

fear & loathing

Death and Resurrection
Death and Taxes
Tweedledum & Tweedledee
Gomorrah (and Sod 'em)

Hmmm... Just as well I don't get involved in host naming.

(Though a friend found that her new Orkplace (-3) had the 12 Macs
named after the signs of the zodiac. As they were upgraded/replaced
the names got changed to scientists (specifically ones who had
worked on the Manhatten Project). Management didn't tumble to it.)

Chris. (Though there were complaints about having to spell Szilard.)

[Anthony de Boer - USEnet]

Post by chris+ (Chris Suslowicz)
Death and Resurrection
Death and Taxes
Tweedledum & Tweedledee
Gomorrah (and Sod 'em)

We had conquest, war, famine, and death on our network at one point.

In fact, other than there being No More War, the other three are still
not decommissioned.

Post by chris+ (Chris Suslowicz)
(Though a friend found that her new Orkplace (-3) had the 12 Macs
named after the signs of the zodiac. As they were upgraded/replaced
the names got changed to scientists (specifically ones who had
worked on the Manhatten Project). Management didn't tumble to it.)

Job[-1] started off with the 12 signs of the zodiac, except that we had
to skip the 4th because the boss smoked.

After that, in fact, we went to philosophers and theologians, many of
the former set being natural philosophers (what you'd call scientists
now). The biggest stretch there was the webserver named Guericke (he
did early research into vacuum) because All Websites Suck.

[Mike Andrews]

Post by Anthony de Boer - USEnet

Post by chris+ (Chris Suslowicz)
Death and Resurrection
Death and Taxes
Tweedledum & Tweedledee
Gomorrah (and Sod 'em)

We had conquest, war, famine, and death on our network at one point.
In fact, other than there being No More War, the other three are still
not decommissioned.

Post by chris+ (Chris Suslowicz)
(Though a friend found that her new Orkplace (-3) had the 12 Macs
named after the signs of the zodiac. As they were upgraded/replaced
the names got changed to scientists (specifically ones who had
worked on the Manhatten Project). Management didn't tumble to it.)

Job[-1] started off with the 12 signs of the zodiac, except that we had
to skip the 4th because the boss smoked.
After that, in fact, we went to philosophers and theologians, many of
the former set being natural philosophers (what you'd call scientists
now). The biggest stretch there was the webserver named Guericke (he
did early research into vacuum) because All Websites Suck.

Hmmmm ... from theologians to sets of theological things.

Faith, hope, and charity.

Pride, avarice, &c.

[Anthony de Boer - USEnet]

Post by Mike Andrews

Post by Anthony de Boer - USEnet
After that, in fact, we went to philosophers and theologians, many of
the former set being natural philosophers (what you'd call scientists
now). The biggest stretch there was the webserver named Guericke (he
did early research into vacuum) because All Websites Suck.

Hmmmm ... from theologians to sets of theological things.
Faith, hope, and charity.
Pride, avarice, &c.

Ah, so you want to name a server after Graham.

Admittedly, the philosophers and theologians scheme started off with a
pair of machines named after the French/Swiss theologian John Calvin
and the English social philosopher Thomas Hobbes.

[Peter H. Coffin]

Post by Mike Andrews
Hmmmm ... from theologians to sets of theological things.
Faith, hope, and charity.
Pride, avarice, &c.

subsection of the house's named file:

...
; the ravens
hunin IN A 192.168.1.11
munin IN A 192.168.1.12
; others
hel IN A 192.168.1.13
hodur IN A 192.168.1.14
ull IN A 192.168.1.15
vali IN A 192.168.1.16
vidar IN A 192.168.1.17
frigga IN A 192.168.1.18
freya IN A 192.168.1.19
balder IN A 192.168.1.20
aegir IN A 192.168.1.21
; *** DHCP ***
; the norn
skuld IN A 192.168.1.32
verthandi IN A 192.168.1.33
urd IN A 192.168.1.34
; the erinyes (furies)
alecto IN A 192.168.1.35
tisiphone IN A 192.168.1.36
magaera IN A 192.168.1.37
; the fates
clotho IN A 192.168.1.38
lachesis IN A 192.168.1.39
...

There's just enough DHCP devices in the house that clotho's usually
assigned, and lachesis usually isn't. About half of the devices on
static IPs (below the ravens numerically) are turned on at a time, but
they're all old enough and historical enough that static addresses make
good sense. One's nutty enough that it can cope with WEP *or* DHCP, but
not both.

For added amusement:

tivo IN CNAME hunin
;nas IN CNAME munin
;fileshare IN CNAME munin

munin hasn't been ordered yet, but the budget's almost there for it.

[Mike Andrews]

Post by Peter H. Coffin

Post by Mike Andrews
Hmmmm ... from theologians to sets of theological things.
Faith, hope, and charity.
Pride, avarice, &c.

tivo IN CNAME hunin
;nas IN CNAME munin
;fileshare IN CNAME munin
munin hasn't been ordered yet, but the budget's almost there for it.

I had thought that the ravens were Hugin(n) and Munin(n), and I see
that various sites agre with me. I'll give the wikipedia reference,
though it's by no means the only one:
<http://en.wikipedia.org/wiki/Hugin_and_Munin>

The way things are going, Tivo and NAS will wind up in a pantheon
fairly soon.

--
"If God had intended us to vote, he'd have given us candidates."

[William R. Walsh]

Hi!

Post by Menno
What is it about hostnames that so gets up people's
snouts?

Hum...after much thinking I dunno.

My employer has been pretty freewheeling about what hostnames machines
on the network were given. Many are purpose or location driven, but
some are just for fun...two of the more interesting ones I've used
were "terrible-lizard" (I'm told it's a rough translation for
"dinosaur") and "dangerous" (this one was given to a PS/2 Model 95
that was set up to pull audit logs from an electronic locking system).
This was primarily because you need a "dangerous" to go with your
"safe" (yes, you can groan, but I don't really care) and also because
that's what the machine would become if I were to become annoyed and
send it sailing across the floor into a helpless but annoying luser's
knees, thusly knocking them to the floor and (hopefully) plowing over
them.

I was only ever asked about "terrible-lizard"...other than that,
nobody has cared. Maybe that's because I haven't yet tried profanity
as a machine's hostname...?

William

[Mike Andrews]

Post by William R. Walsh
Hi!

Post by Menno
What is it about hostnames that so gets up people's
snouts?

Hum...after much thinking I dunno.
My employer has been pretty freewheeling about what hostnames machines
on the network were given. Many are purpose or location driven, but
some are just for fun...two of the more interesting ones I've used
were "terrible-lizard" (I'm told it's a rough translation for
"dinosaur") and "dangerous" (this one was given to a PS/2 Model 95
that was set up to pull audit logs from an electronic locking system).
This was primarily because you need a "dangerous" to go with your
"safe" (yes, you can groan, but I don't really care) and also because
that's what the machine would become if I were to become annoyed and
send it sailing across the floor into a helpless but annoying luser's
knees, thusly knocking them to the floor and (hopefully) plowing over
them.
I was only ever asked about "terrible-lizard"...other than that,
nobody has cared. Maybe that's because I haven't yet tried profanity
as a machine's hostname...?

I think that profanity and vulgarity would be entirely appropriate.
That'd give me hostnames like Maricon, Huevos, Mierda, Merde, Scheiss,
Bakayaro, Kuonyaro, Chyort, ... . You get the idea.

Oh, and those would be names assigned to user-community boxen. The
swervers would have names from a somewhat more restrained subset of
the dictionary.

[Richard Johnson]

Post by Mike Andrews
I think that profanity and vulgarity would be entirely appropriate.
That'd give me hostnames like Maricon, Huevos, Mierda, Merde, Scheiss,
Bakayaro, Kuonyaro, Chyort, ... . You get the idea.

$ORK accidentally edged into that territory, by reference. The big
boxes were named after Native American tribes like arapahoe, kiowa,
cheyenne, and so forth. This of course led into geographic place names
like arapahoe, kiowa, and so forth (Indian Peaks Range in Colorado).

Naturally, other features had to be used for the file server names. They
started with northpark. When northpark was retired, it was replaced by
middlepark. Then, when middlepark was retired, Cartman's cussin' had
manglement rather worried, so the replacmement was named... fileserver.

As often as it dies, it should have been named kenny.


Richard

[Satya]

Post by Richard Johnson
As often as it dies, it should have been named kenny.

Our windows box (for the purpose of rdesktoping to and saying "see, it
works fine in *our* non-malware infested IE") is named kenny.

Oh goddammit, it's named kyle. kenny was a different box.

[Howard S Shubs]

Post by Satya
Oh goddammit, it's named kyle. kenny was a different box.

So Kenny died?

[Satya]

Post by Howard S Shubs

Post by Satya
Oh goddammit, it's named kyle. kenny was a different box.

So Kenny died?

I think so. Not sure what happened to it.

[Howard S Shubs]

Post by Satya

Post by Howard S Shubs

Post by Satya
Oh goddammit, it's named kyle. kenny was a different box.

So Kenny died?

I think so. Not sure what happened to it.

Too many antacids, prolly.

[Joe Zeff]

On Sat, 24 Jun 2006 00:40:49 GMT, William R. Walsh

Post by William R. Walsh
I was only ever asked about "terrible-lizard"...other than that,
nobody has cared. Maybe that's because I haven't yet tried profanity
as a machine's hostname...?

Back when I first started dual-booting into Yvahk, it wanted a host
name. At that point, I was orking for $WEPUSHPACKETS, and their
scheme was countries. Not wanting to overlap, but trying to fit in, I
decided on khorlia, an imaginary country the late Dave Fox used to
write about. I've decided that if I ever have my own domain, here,
I'll use places you can't visit, such as freedonia, klopstokia,
concordia and so forth.

[Kevin @kevingoebel.dot. com>]

On Sat, 24 Jun 2006 18:43:07 GMT, Joe Zeff

Post by Joe Zeff
On Sat, 24 Jun 2006 00:40:49 GMT, William R. Walsh

Post by William R. Walsh
I was only ever asked about "terrible-lizard"...other than that,
nobody has cared. Maybe that's because I haven't yet tried profanity
as a machine's hostname...?

Back when I first started dual-booting into Yvahk, it wanted a host
name. At that point, I was orking for $WEPUSHPACKETS, and their
scheme was countries. Not wanting to overlap, but trying to fit in, I
decided on khorlia, an imaginary country the late Dave Fox used to
write about. I've decided that if I ever have my own domain, here,
I'll use places you can't visit, such as freedonia, klopstokia,
concordia and so forth.

Don't forget to name one Panavision... they film a lot of movies there and
the scenery is marvelous.

Kevin

[Menno]

On Sat, 24 Jun 2006 18:43:07 UTC, Joe Zeff

Post by Joe Zeff
On Sat, 24 Jun 2006 00:40:49 GMT, William R. Walsh

Post by William R. Walsh
I was only ever asked about "terrible-lizard"...other than that,
nobody has cared. Maybe that's because I haven't yet tried profanity
as a machine's hostname...?

Back when I first started dual-booting into Yvahk, it wanted a host
name. At that point, I was orking for $WEPUSHPACKETS, and their
scheme was countries. Not wanting to overlap, but trying to fit in, I
decided on khorlia, an imaginary country the late Dave Fox used to
write about. I've decided that if I ever have my own domain, here,
I'll use places you can't visit, such as freedonia, klopstokia,
concordia and so forth.

Heh! My T/D domain is called "utopia".

[Paul Kelleher]

In article <K9l6boYvJbtS-pn2-***@localhost>, ***@xs4all.nl
says...

Post by Menno
On Sat, 24 Jun 2006 18:43:07 UTC, Joe Zeff

Post by Joe Zeff
I've decided that if I ever have my own domain, here,
I'll use places you can't visit, such as freedonia, klopstokia,
concordia and so forth.

Heh! My T/D domain is called "utopia".

Hmm. Utopia being unvisitable. Sounds about right.

Our HQ has a naming scheme that they're _so_ proud of, they want to
inflict^Winfect^Wforce us all to use. Me, I'm quite happy with our
site's mail server being 'mail' and our database being 'database'. We
won't ever need two of either, and out budget wouldn't stretch to it
anyway. We already have failover systems; they just don't have anywhere
to failover to. It doesn't stop them.

HQ demands^Wrequires^Wcan't see the problem with us having machines
called NTSWX01, NTSWX02, LXSWX03, NTWWX04, 2KSWX05, and so on. Thus
neatly removing the need for a DNS handler, and instead using a small
Perl script. Except no, we can't do _that_. They've installed (for us,
because it's too technical to do ourselves) the W32 build of ISC BIND.
On the most unstable, lowly-specced machine we've got, which is doing
double duty as an NT PDC, and is the point on our network where all
the other networks in the group have their trust relationships with,
so we can't just up and reformat-reinstall it. That's NTSWX01, because
PDC was too long a name.

Now, was that WreXham NT Server 01 I should have rebooted? It's so
hard to tell. I'm sure there's someting in the relevant RFC about not
picking random or numerical hostnames, except the relevant RFC Wasn't
Invented There...

AHS.[1]

Kelloggs

[1] I could also go with 'hierarchies' or 'HQs' here. Hieraches[2].
[2] Or, indeed, hierAIX. They've inflicted^Wordered that upon us, too.

[Paul Arthur]

Post by Menno
What is it about hostnames that so gets up people's snouts? I've seen
horrid constructions. There's the school of thought that thinks it
would be a good idea to stuff as much information as possible into the
hostname, leading to names like "rs6000-aix51-2002-03-01-1". (Which of
course runs AIX 5.3, because renaming it after the upgrade would have
been Difficult). Then, there is the "We need to know that it belongs
to Our Company" school of thought, leading to
MyCompany0001.mycompany.com. Or the "Lack of imagination" school of
thought: "box0001".

Ah. At Ork we have ustroyfs06a. Which is, naturally, a fileserver located
in Troy. Quite possibly the sixth one (buggered if I know what the 'a'
is for, though). And, of course, there's usfstryoa3, which is the third
fileserver located in Troy. The others, which I thankfully cannot
recall at the moment, follow one of the patterns above, except for when
they don't.

Also worth a laugh is the naming convention for the *nix servers.
oak001a, pine004b, balsa002a, (...), chestnut, and elfin.

Post by Menno
Anyways, everything is working again. Time for beer and merriment.

Mmm, merriment. Where do you buy that, again?

[Menno]

On Sat, 24 Jun 2006 03:14:10 UTC, Paul Arthur

Post by Paul Arthur

Post by Menno
Anyways, everything is working again. Time for beer and merriment.

Mmm, merriment. Where do you buy that, again?

's a flavour of crisps.