Trip to the Land of Twinkies, Day 4

So, I've been kind of light on content this week.

Part of this is because I made the decision to leave my camera in a safe in College Park. This was a wise choice from a logistical standpoint, as I had to walk over 2 miles from the West Natick MBTA commuter rail station to the Hampton Inn.

That's fine - I haven't been able to go anywhere, and the only thing I've wanted to take a picture of are the bright red trees out in the parking lot. Little maples, very pretty.

Anyway, I was planning on doing some google maps editing and uploading my GPS log from the trip up here. We hit 151 miles an hour at one point, and apparently the whole notion of lining out the train route up the Eastern excited me. Or something.

Unfortunately, here's what I've been dealing with since I got here Monday night:

Macintosh-4:~ natkuhn$ ping 128.8.76.2
PING 128.8.76.2 (128.8.76.2): 56 data bytes
64 bytes from 128.8.76.2: icmp_seq=0 ttl=47 time=203.091 ms
64 bytes from 128.8.76.2: icmp_seq=1 ttl=47 time=999.273 ms
64 bytes from 128.8.76.2: icmp_seq=2 ttl=47 time=394.044 ms
64 bytes from 128.8.76.2: icmp_seq=3 ttl=47 time=839.724 ms
64 bytes from 128.8.76.2: icmp_seq=4 ttl=47 time=419.128 ms
64 bytes from 128.8.76.2: icmp_seq=5 ttl=47 time=1003.461 ms
64 bytes from 128.8.76.2: icmp_seq=6 ttl=47 time=1004.051 ms
64 bytes from 128.8.76.2: icmp_seq=7 ttl=47 time=1003.662 ms
64 bytes from 128.8.76.2: icmp_seq=8 ttl=47 time=954.812 ms
64 bytes from 128.8.76.2: icmp_seq=9 ttl=47 time=1010.508 ms
64 bytes from 128.8.76.2: icmp_seq=10 ttl=47 time=1003.984 ms
64 bytes from 128.8.76.2: icmp_seq=11 ttl=47 time=1003.854 ms
64 bytes from 128.8.76.2: icmp_seq=12 ttl=47 time=1003.493 ms
64 bytes from 128.8.76.2: icmp_seq=13 ttl=47 time=1003.364 ms
64 bytes from 128.8.76.2: icmp_seq=14 ttl=47 time=1003.827 ms
64 bytes from 128.8.76.2: icmp_seq=15 ttl=47 time=1003.112 ms
64 bytes from 128.8.76.2: icmp_seq=16 ttl=47 time=1008.885 ms
64 bytes from 128.8.76.2: icmp_seq=17 ttl=47 time=1002.986 ms
^C
--- 128.8.76.2 ping statistics ---
19 packets transmitted, 18 packets received, 5% packet loss
round-trip min/avg/max/stddev = 203.091/881.403/1010.508/248.835 ms

The values <1000ms are cool; the internet works when that happens. Even the values > 1010ms are generally cool; I've had working internet here when it gets up to 2200ms - working, but slow.

It's all those values - sometimes hundreds in a row when I have non-working internet - that hit between 1002 and 1004 that are convincing me there's some kind of artificial packet shaping at work here, and that it's horribly broken.

I tried to find the Hampton Inn tech support contact number, and instead, I found installation/configuration guide PDFs meant to be used by contractors installing this stuff in all kinds of hotels. In them, it says that Hampton Inn (and its parent, Hilton) asks the contractors to prevent contact between user computers.

Hmm... bonjour, anyone?

I turned off mDNSResponder and mDNSResponderHelper using launchctl and actually got working internet. It's been worse today, but still up more than down. (of course, as I type this, I appear to be stuck in a good stretch, with over 60 pings <30ms.)

Anyway, the situation is affecting all web traffic, including Hampton Inn websites, and all traffic associated with IMs, and most traffic associated with IMAP. SMTP is out completely when this happens.

Well, as it is, FileMaker training is going well. I've learned a lot of optimization tips for the database work I've done over the years, and also learned that nothing I was doing was outright incorrect in the professional sense - they live by the rule "if it works, and it does what you want, it's right." Today was calculations, tomorrow will be scripting. I've spent a lot of time in these, so it's actually been easier for me to deal with than the database planning stuff we did yesterday and Tuesday. An aside - Did you know that FileMaker has average and mode functions, but not median? We learned how to reinvent that wheel today.

So yeah, that's the story. Sorry for the blank slate. Again. Or as always.

Edit:: as if that weren't bad enough, it would appear there is a configuration error on our xserve. Now, my website is ugly. This is because everything lives on that xserve.