Hi all,

I admit this has been a particularly bad year for me, but it isn't THAT unusual by my standards. I typically spend I would reckon quite honestly about 70-80% of my time with computers just fixing stuff up that either I broke or was broke in the first place. Usually it's the latter by some considerable margin. That's really inefficient, isn't it? I just wondered what others' experiences are of this phenomenon. Obviously I don't do computers for a living or I would have starved years ago.

So what is the Panel's consensus view on the percentage of wasted time we should expect to spend, on average, "running just to stand still" on maintenance tasks?

CC.

As a consultant I would estimate the vast majority of my time is fixing other peoples stuff, I spent probably an hour a week (usually package maintenance) working on my gear unless I'm installing something new and then the time varies greatly, I probably spent 5-6 hours on installing my ASA by the time I had the tunnels and everything setup as I wanted. I probably spent ~20-40 hours installing my linux machine that functions as a game, teamspeak, web, and email server for various people... maintenance on either is negligible though once it's in place. If you're spending 70-80% of the time repairing something broken, it either wasn't setup in the first place correctly or you may be tinkering too much with the stuff you have in place... and need a dev env for testing.

I personally make the mistake of getting curious about 'trying something new' and end up doing a good amount of damage to my systems in the process.

Taking five or ten minute breaks while working on a problem often helps give me perspective, that I don't 'really' need to be working on whatever issue I'm facing; that there are often simpler solutions than the one I'm using if I'd really think on it. I have to remember to stop being so stubborn, and just get whatever I need to do, done.

I try to make good images. I don't try to fix, I simply reload.

Fair enough. But how can you be sure your images are good before the fateful time when you might actually need to rely on them?

Presently, I'm experimenting with using a software RAID array, to see how secure my data is on it. So far, I've been surprisingly pleased. Once in place, it seems to be humming along, quite fault tolerant, and faster than I expected.

I've also learned (the hard way) that it's not enough just to do back ups; one must back up *effectively.* I have an unusual back up system, that I've never seen used by others. On every system, I install a small 'recovery' system. It's a minimal installation of whichever OS I'm using (Linux & OSX at present) that will enable me to dual boot into that recovery system. From there, I can execute a full, offline backup of my primary system before attempting any sort of major upgrade or configuration. This not only ensures that I have a solid image, but several solid images at various states of my system's lifespan. More importantly, it gives me a chance to double check that the backup server is still functional.