I know this would be a pretty radical change, but it would eliminate all possability of unexpected/unwanted sector writes, and would serve as a fail-safe for when we muck things up.
Something like IOS’s “copy run start” is what I’m thinking of.
Yes, having a running config and a saved config would be good, particularly when you’re remotely administering and you accidently kick yourself out, a reboot solves the problem.
Safe mode works fine as a fail-safe (and I’m glad to have it, it’s saved my butt more than a few times), if you are using a command line (which I prefer, but most others at my company always use WinBox).
Also it does nothing to prevent flash writes, such as those that seem to occur every time a PPPoE login/logout creates/deletes a Queue/Mangle rule. Or when I export/backup to a file, I always retreive it right away with FTP (or have the MT mail it to me), and have no reason to keep it around on the MT after that. There also many times I want to make a temporary change, which will last longer than a terminal session, but not persist across reboots.
Not individualy, no. But over time, with enough writes, Flash memory does go south.
I have only had it happen once, due to a rather stupid configuration mistake on my part (cache to flash, bad), but after ~13,000,000 sector wites, it wouldn’t take any more.
I also have a couple of other MT boxes, where the count is well over 1,000,000. I haven’t had issues yet, but I’m concerned about what state they will be in a year from now.
Any cheap hard drive has about the same life expectancy… if you spent a few bucks on a maxtor or other you will only get 1-2 years out of it Why do you think they only have a 1 year warranty now?