COMPUTER CRASH POST-MORTEM
or
Don’t do what I did

On Tuesday I arrived at the library to send already-composed emails and make a blog post. The laptop wouldn’t boot.

Bear Bussjaeger quickly and correctly diagnosed a hard drive problem; the drive was going south, boot sector first.

But the real problem wasn’t a dramatic crash of the hard drive. It was more like a slow leak between my ears. This post is to recount where I went wrong, note the very few things I did right, thank the people who saved my backside, and serve as a cautionary tale for anybody out there who might be stumbling into similar screwups.

—–

Where things stand now: It’s been three days and the trouble is far from over. The fact that I have even limited online access and email contacts with a few close associates is entirely due to the standard little LOT of help from my friends.

Big IOUs to:

• S for providing the iPhone that I didn’t really want, but that turned out to be a lifeline; and for setting me up with a new phone-based email address when nothing else was working, then establishing contacts with the tech guys;
• To tech guys B & B for getting me out of password hell;
• To M & J, who both rushed to offer lightly used computers;
• To D, who suggested getting SpinRite to try to salvage the drive, and M who is loaning me a copy.

Without these people I would be so screwed I doubt I’d have managed to post or make contact with anybody for weeks.

As is, I still have a long trek ahead to have a fully functional computer, and at this point, it’s still entirely possible that I could end up losing a lot of mail plus the ability ever again to decrypt and read the last year’s worth of encrypted messages and data.

I’m optimistic SpinRite will keep that from happening, but who knows?

—–

Where I went wrong: I’m tempted to say “everywhere.” But my mistakes fall into two broad, and sadly familiar categories, both of which apply to all kinds of prepping, well beyond the realm of computers:

• I got complacent.
• I didn’t think out my recovery plans.

Anybody else recognize themselves there?

Specifically, I got into the mindset that this “new” computer worked so well that it would just go cranking along forever. I unthinkingly assumed that — even though the “new” computer was actually nine years old, had a keyboard with half the letters worn off, and had been giving out weird glitches for months. Oh, the wages of optimism.

And I got so much into the habit of treating key data in certain rote ways that I long since stopped asking questions like, “What will I do if X, Y, or Z happens?”

The specific problems that arose from those two broad follies were many and varied.

—–

System backups. I habitually backed up documents and valuable data like passwords, email server settings, and bank information, but I stopped doing system backups. Even if I’d been doing system backups, I probably wouldn’t have given enough thought to questions like, “How do I apply these backups to get up and running again without delay?”

Encryption. I send and receive a lot of PGP or GnuPG encrypted email. All my most private data I store encrypted. So then why (you might ask) did I not a) back up my encryption keyrings after creating a new key pair a year or so ago and 2) make sure I had a working encryption system somewhere apart from my everyday computer? Instead, I ended up with “safely” encrypted vital data that I can’t decrypt.

A functional backup computer. Given my reliance on the computer to earn a living and communicate with the world, I should not only have had backups, but should have had a fully functional, frequently tested second computer with those backups on it, up-to-date and working. As is, the old computers I thought I was so smart to keep on hand were, and are, completely out of date and in various stages of malfunction. I can’t use them at all to get back online or find current, much-needed data.

Contact information. Although in a way this is a subset of the other problems, I should have saved my email address book and other contact information separately and in the clear (unencrypted).

Alternate email setups. Another subset, but as it turned out, a huge one. Having lost access to my email accounts AND the (encrypted and unavailable) means of setting them up again quickly, I was skunked. I couldn’t communicate with people I desperately needed to reach and I couldn’t receive password hints, verification codes, etc. from sites. Of all the problems I caused myself, that was the worst. That non-existent functional backup computer should have been running a fully parallel email system. (And no, don’t say web mail; I hate web mail.) Point in my favor: I had parallel email on the iPhone. How could I know Apple would “upgrade” it to death AND that my main email servers would be having trouble this week? “Two is one and one is none” and sometimes five or six can also be none.

Security missteps. My otherwise positive stand on privacy came back to bite me in several ways. I’ve refused to use a password manager because I don’t trust anybody but myself to create and keep my passwords. I just kept them in a form I momentarily can’t access. I’ve also habitually refused to give a mobile number to most sites, and that meant that, without working email, some sites had no way to verify my identity when I entered them via an unfamiliar computer at the library.

Passwords and “remember me.” The passwords I always manually enter (e.g. on bank sites) I had no problem using, post-crash, on the library’s desktop computers. The ones on which I allowed browsers to “remember me” caused trouble because I didn’t have to remember them for myself — until I did have to. And couldn’t. Nor could I receive password hints or password resets from those electronically “remembered” sites.

Definitely a series of unfortunate events.

So much pure stupidity. Laziness. Failure to plan. Cheery complacency. And all this is especially bad coming from me, the preparedness believer. I confess it here because my folly might help somebody else prevent a similar mess.

I shudder when I think of the fix I’d have been in without that handful of helpful, on-the-ball, generous friends.

—–

Did I do anything right? Not much.

Partitions. I set up the computer so that my created data and my settings were on two different partitions, separate from the operating system. While that doesn’t seem to have made a huge difference here, it does make it harder for a crash to destroy everything. And it is easier to back up those partitions even while failing to back up the system.

The operating system itself. Mint Linux. In this case version 19.1 Cinnamon. User friendly and gives lots of control. I can re-install it easily myself if it goes haywire, and I can install it easily on another computer without worrying about licensing fees, permissions, or nannying from outside forces who believe that they, not I, should control my machines. The problem in this case was not the OS, but having such an easy-to-work-with operating system meant I could quickly verify that this was a hardware issue, not a software issue.

Encryption. Well, anyhow, it would have been a good thing, had I had an updated, parallel system running on another good quality, fully functional system.

—–

So there it is; a classic example of “do as I say, not do as I do.”

To sum (and a memo to myself):

• Do full system backups. Do them frequently and on a regular schedule.
• Have a parallel system designed to swing right into action if the primary fails.
• If you encrypt, back up your keys and make sure your parallel system is also fully ready to encrypt and decrypt.
• Keep key contact information where you can access it.
• Think out various contingency plans before you need them.

—–

Now I’m sure that readers more tech-savvy than I will have lots to add, particularly about ways and tools to make all this (and more) workable.