On Thu, 07 Apr 2005 15:45:02 GMT, Doug Miller wrote:
In article , Dave Hinz wrote:
On Thu, 07 Apr 2005 09:41:30 -0400, Paul Kierstead wrote:

That is one nice set-up, Dave. I don't like my bandwidth providor much,
don't use their email services and their DNS used to drop all the time,

Does the name sound at all like "voyager.net" by any chance?

Oh, geez, don't get me started on Voyager. For years, I had an account with a
local ISP. Hardly ever had any technical problems, and when I did, the
customer service was superb.

That would be execpc, yes?

Then the local outfit was bought ought by
Voyager, and almost immediately everything went in the toilet. Dropped
connections, dial-up modems that wouldn't answer, sloooooow connections, at
least fifty times the amount of spam... and absolutely clueless customer
service.

Yup.

The last straw was when I called for about the seventeenth time to
complain about the slow connections. They insisted it had to be a problem in
my configuration (which hadn't changed for six months). Told me to run
traceroute to show where it's getting bogged down. OK... connection between me
and them was just fine. Two hops farther, though, it was a *very* different
story - repeated timeouts between two of *their* machines.

The droids don't like it when the caller understands more than they do,
do they.

The customer
disservice rep told me to reinstall Windows. Yeah, right. I told him, sure,
I'll do that, just as soon as you explain to me how anything on *my* machine
can cause packet timeouts between *your* servers.

Heh. Now try being the server design guy for an e-commerce site (ahem),
who gets a ****ed off customer call escalated to him (why me? because I
have this "Sherlock Holmes" reputation or something). So, angry customer
says "I can't reach your site, you people aqre the suxx0rs!!@@!!!!!!".

OK, let's see. Go to this checklist-diags site I wrote. Can you see (item)?
Yes? How about the big red star? No? OK, that's interesting. Click on
"test 3" link please. No? I see. No, I'm sure our site is up, we've
processed 3000 transactions this morning already. OK, run a traceroute,
like this...

Ah, that helps. The problem is on that fifth hop there. The bad news is,
it's on equipment not owned by either you, or me, or your ISP, or my
ISP. It's actually equipment owned by someone in the middle, and neither
you nor I are their customer. Which means that we're both going to be
equally unable to get them to do anything, but we'll try since you're the
customer. Yes, sorry, that's the best I can do. No, I can't "just fix it",
it's not my equipment, and the people who own it don't care about your or
my business. Yes, that's the way the internet works, it's a network.

And then, hours later "No, I didn't change anything, but apparently
(sprintnet, let's say) noticed and fixed their problem."

Sometimes you just can't fix the stuff, but if the droid couldn't recognize
that it was their own switchgear where the traceroute was dying, well,
it's hard to get good helldesk people.