Today I found myself thinking of Terminator 3, specifically the plotline in which all kinds of random computer crashes are spreading across the internet.

For obvious reasons.

In today’s real world incident, it’s a bug in an auto-pushed update for widely-used security software by CrowdStrike, ironically used to protect mission-critical systems. In the two-decade-old movie (pardon me while I turn to dust), it’s Skynet spreading itself across the internet.

At the time, I thought the nuclear strike would wipe out a lot of internet infrastructure, destroying major nodes and leaving pieces of Skynet disconnected from each other. A commenter remarked that he’d been doing research for a novel and experts agreed that enough of the major nodes and infrastructure would survive the attack to keep the network functioning.

The interesting thing: Neither of us had heard the story that ARPANET (the internet’s predecessor) had been designed for that scenario. These days, it’s pretty much repeated as gospel… but apparently it wasn’t a design goal, and the idea that it was can be traced back to a 1991 article in Network World magazine that conflated ARPANET with a different network design, which was never actually built. (via)

From there it took on a life of its own for the same reason many urban legends (and conspiracy theories) do: it made a better story.

Two hot takes (so to speak):

Auto-updaters shouldn’t run when the system is really busy.

And installers that check to see the whether the same or newer version is “already installed” should either be really thorough about what they’re checking, or offer to do a repair install anyway.

Overheat!

I’d fired up a game of No Man’s Sky, which even after redoing the thermal paste and adding another case fan still pushes the limits of my system’s cooling, especially if I forget to wait for all the background processes to finish loading on Windows startup. I left a space station, landed on a planet, started mucking around with the structure I had found, and hit a waypoint pole to save again—

And the system shut down.

I turned it back on to get the fans running again and help cool off. Which worked. But when I logged in, I had a bunch of errors with Google Drive. Apparently its auto-update launched during my GPU-intensive game and overheated it enough to reach the shutdown threshold.

Fortunately, I’d just saved my game — or so I thought.

I spent about half an hour trying to fix Google Drive before I decided to just go back to the game for now.

No Man’s Sky picked up seamlessly…at the previous save. Maybe the write cache hadn’t been flushed yet or something. I’m glad I only lost about a minute of gameplay, though… I’ve got several hundred hours on this save file, and I’d hate for it to get lost or corrupted!

Anyway, back to Google Drive. I couldn’t reinstall it because I couldn’t uninstall it, and despite my efforts I couldn’t remove enough traces of it for the installer to be willing to run. You can read the whole reinstalling Drive saga on my troubleshooting site, along with the taskbar that lost its icons halfway through.

I eventually fixed it by copying the installed program files from another system and running the uninstaller manually. That resolved both the taskbar icons and the Drive installer being willing to run.

Then I dragged myself into bed.

Yeah, it was a fun Saturday night. 🙄

Comic-Con 2011 ticket sales crashed under heavy load shortly after going online.

I think we’re seeing another shift in the process of getting to Comic-Con.

It used to be that, as long as you were aware of the onsale dates and could both plan your trip and pay for your tickets far enough ahead of time, getting those tickets wasn’t a problem. Sure, the show might sell out months ahead of time, but it would take weeks or months to get to that point.

Now, people are looking at preview night already being sold out, and looking back at their last experience with hotel reservations, and freaking out: We can’t just buy our tickets this week – we have to buy them NOW, as soon as they go on sale, or we won’t be able to get in!

Get enough people reacting that way, and it becomes a self-fulfilling prophecy. Fear of a rush ends up creating a rush, just like fear of a run on a bank often triggers one. Add in the live progress bars as a feedback mechanism, and it snowballs even faster.

Until it crashes the server, anyway…

(Originally posted as a comment at The Beat.)

A few weeks ago, I stopped at the post office on the way to work. As I walked to the door, my eye was immediately drawn to the big hole in the wall where, apparently, a car had crashed into the building while trying to park.

Whatever vehicle had done the damage was long gone, and the area around was cordoned off with yellow caution tape. Since I like to take pictures of weird stuff, I snapped a photo with my cell phone. (No, I didn’t cross the tape; it’s just not visible in the frame.)

This morning, I stopped at the same post office again. This time, my eye was drawn to a new addition to the facility:

One of these bright yellow posts stood in front of each parking space along the building. You can still see where the concrete has been spread around the base, and absolutely no paint has worn off. Clearly someone decided not to take any more chances with wayward cars!

I put the Pushing Daisies soundtrack CD in my computer and it rebooted. Fortunately it took less than a minute….

Of course I tried it again to see if it was a fluke. Same thing. Then I decided to try with another CD. Apparently any audio CD crashes it. Shows how often I use CDs these days.

At least the DVD-RW drive works. IIRC it doesn’t have CD-audio hooked up, but the player seems perfectly happy to just read it digitally.

I guess I’ll have to set aside some time on Sunday to figure out what’s causing the problem. It’s a bit late tonight, and tomorrow’s going to be busy!

First, some linkblogging…

And then the “fun” started.

  • Me: I’m going to focus on project X today!
    Computer needed for project X: I’m going to lock up today!
    Me: Argh!
  • Someone thought it would be a good idea to cover “Wonderful Christmastime?” 😯 (For the record: Shazam says it was Hilary Duff.)
  • OK, after 3½ hours stuck at 74%, I think I can assume chkdsk is stalled. *grumble*
  • Ate some blackberries I’d forgotten about from a week ago. Good news: I only threw out 1! Bad news: I should’ve thrown out 2. Blech!
  • Chkdsk round 3 is at 55% on Stage 5 of 5. Going to call it a night & hope my PC runs tomorrow. (According to Facebook, this posted at 5:55pm!)

Computer update: Disk check finished overnight, seems OK. Ran a backup just in case, but got some work done on that project!

The IEBlog recently posted about their efforts to improve reliability in Internet Explorer 8, particularly the idea of “loosely-coupled IE” (or LCIE). The short explanation is that each tab runs in its own process, so if a web page causes the browser to crash, only that tab crashes — not the whole thing. (It is a bit more complicated, but that’s the principle.) Combine that with session recovery (load with the same set of web pages, if possible with the form data you hadn’t quite finished typing in), and you massively reduce the pain of browser crashes.

I’d like to see something like this picked up by Firefox and Opera as well. They both have crash recovery already, but it still means restoring the entire session. If you have 20 tabs open, it’s great that you don’t have to hunt them down again. But it also means you have to wait for 20 pages to load simultaneously. It would be much nicer to only have to wait for one (or, if I read the IE8 article correctly, three).

Edited to add:

On a related note, I’ve run into an interesting conflict between crash recovery and WordPress’ auto-save feature. If you start a new post, WordPress will automatically save it as a draft. If the browser crashes, it will bring up the new-post page, but restore most of the form data you filled in. So the title, the text of your post, etc will all be there. But WordPress will see it as a new post, and you’ll end up with a duplicate.

This wasn’t a major problem when I encountered it — I had to reset the categories, tags, and post slug after I hit publish (since I hadn’t noticed that they’d been reset to defaults), and I just deleted the older, partial version of the post — but I can imagine if I’d uploaded an image gallery, I would have been rather annoyed, since there’s no way (that I’ve noticed) to move images from one post to another. Reuse them, sure, but not such that the gallery feature would work.