Re: Uptime records

["George Georgalis" <george@xxxxxxxxx>]

On Mon, Apr 11, 2005 at 08:41:24PM -0000, Andreas Hauser wrote:
>phessler wrote @ Mon, 11 Apr 2005 12:28:47 -0700:
>
>> http://marc.theaimsgroup.com/?l=openbsd-misc&m=101480739621320&w=2
>> 
>> :D
>
>Sure but i'm more interested in uhm real time :P
>
>You can fake speed benchmarks too, but that won't change the reality that
>openbsd is slow and theo doesn't care about it.
>
>And 187 days is far from my goal or what i have reached with other systems.
>But then 2400 days ago there was no DragonFly BSD ...

I think more important than big uptime, is scheduled
maintenance. I've often pondered the idea of a progie
that captures the uptime and increments a reboot
counter for (and only for) scheduled power cycles.
guess it could be accumulated in a file, but never
seemed worth the trouble.

I can say that uptime roll-overs suck, spent half hour
researching why a Linux box rebooted one day, turns out
to have been on 2**16 seconds, blah. There is a pretty
good bug report for it... my client likes his uptime
so I came up with this hack.

echo $((`cat /proc/uptime | sed 's/\..*/ /'` + 42949673 )) | sed 's/$/ seconds/' | secs2dhms 
578 days 23 hours 56 mins 20 secs

What's the dfly rollover?

// George


-- 
George Georgalis, systems architect, administrator Linux BSD IXOYE
http://galis.org/george/ cell:646-331-2027 mailto:george@xxxxxxxxx