Slashdot Mirror
Ask Slashdot: How Much Did Your Biggest Tech Mistake Cost?
NotQuiteReal writes: What is the most expensive piece of hardware you broke (I fried a $2500 disk drive once, back when 400MB was $2500) or what software bug did you let slip that caused damage? (No comment on the details — but about $20K cost to a client.) Did you lose your job over it? If you worked on the Mars probe that crashed, please try not to be the First Post, that would scare off too many people!
DNS servers on the same subnet. You, know, the thing you aren't supposed to do, but everyone does anyway.
Pretty much all modern Intel CPUs from the past many years.
Be careful about criticizing others. Routers don't have default gateways, they have null routes. They can also be set up to be redundant gateways for others and have many redundant null routes themselves...
Turning off STP on just one router would never be a problem. There are master and standby root bridges. Even if they both go down, others will step in to take the job. It would require a total network shutdown of all layer three equipment before it would be a problem and even then, ttl limits and excess traffic would cause the routers to drop one of the cables in the loop within seconds.
This is entry-level networking knowledge.
Databases should be backed up with a text-dump (such as an SQL INSERT list), not the actual database file, because of the internal pointers that are fragile. A text-dump "flattens" the pointers. If you do use the actual database file as a backup, shut all DB writing off first, during the backup. And keep multiple generations.
Table-ized A.I.
The poster was not the boss. The boss calls the final shots. The technician's job is to present the risks (trade-offs) as accurately and clearly as possible. If the boss(es) then choose to ignore the risk warnings, the blame falls on them. If you usurp their power, you are out the door (unless it's a legal matter).
Incidentally, I was in a somewhat similar situation where marketing planned to release about 30 websites for satellite offices all at once along with a press release about the new sites. I pointed out our "budget-oriented" infrastructure may not be able to handle such a sudden load, and suggested staggering the releases. Other technicians agreed with my warning, but the marketing chief was really disappointed, saying something like, "It's better P/R to have one big release. Staggering the releases takes the punch out of it."
I was tempted to respond, "30 crashed sites is not good P/R either", but smartly bit my tongue (based on prior experience with "reality" statements). He was a true P-H-B, always looking for a cheap short-sighted shortcut, but tried to blame us when his paper tigers got eaten. He drove one guy to retire early. Later he was under investigation for giving contracts to his buddies instead of basing them on merit. Not surprising, his buddies were also idiots.
Table-ized A.I.
Heating it up causes the metal to expand which can unjam a stuck head in some circumstances.