Hi my primary dns server was up for 500 days. Now I see in top that it's been up for only 3 days. No trace of reboot neither in messages nor in last log. Does top start over at 0 if a certain amount of up-days have been reached? Or how do I have to interprete that? thanks a lot Philipp
On Sep 8, mailinglists
my primary dns server was up for 500 days. Now I see in top that it's been up for only 3 days. No trace of reboot neither in messages nor in last log. Does top start over at 0 if a certain amount of up-days have been reached? Or how do I have to interprete that? I think there is an overflow at around 480 days. Look at /proc/uptime, maybe there is another value there.
Markus -- __________________ /"\ Markus Gaugusch \ / ASCII Ribbon Campaign markus@gaugusch.at X Against HTML Mail / \
On Monday 08 Sep 2003 11:59, Markus Gaugusch wrote:
On Sep 8, mailinglists
wrote: my primary dns server was up for 500 days. Now I see in top that it's been up for only 3 days. No trace of reboot neither in messages nor in last log. Does top start over at 0 if a certain amount of up-days have been reached? Or how do I have to interprete that?
I think there is an overflow at around 480 days. Look at /proc/uptime, maybe there is another value there.
There has been some discussion of uptime wraparound in summary articles of kernel development on http://lwn.net, IIRC some fixes went in last year, so you could check the exact max. reportable uptime for different kernel versions by googling for it. Rob
On Mon, 8 Sep 2003, Markus Gaugusch wrote:
On Sep 8, mailinglists
wrote: my primary dns server was up for 500 days. Now I see in top that it's been up for only 3 days. No trace of reboot neither in messages nor in last log. Does top start over at 0 if a certain amount of up-days have been reached? Or how do I have to interprete that? I think there is an overflow at around 480 days. Look at /proc/uptime, maybe there is another value there.
Linux counts time in jiffies (usually 1/100 sec on kernel 2.4, 1/1024 sec on 2.6), and on a 32bit system this counter wraps around after 42949672.96 sec or in other words: 497 days, 2 hours, 27 minutes and 52.96 seconds. This 'problem' is tackled in the 2.6 kernel (where it happens after about 50 days), but it goes away as soon as you use a 64bit system... BTW: there is a slight chance that the computer might crash during the 10msec when jiffies==0, so you were a bit lucky that your server is still up and running. Ciao, Roland +---------------------------+-------------------------+ | TU Muenchen | | | Physik-Department E18 | Raum 3558 | | James-Franck-Str. | Telefon 089/289-12592 | | 85747 Garching | Telefax 089/289-12570 | +---------------------------+-------------------------+
Hello All, Monday, September 8, 2003, 12:48:55 PM, Phillipp wrote: m> Hi m> my primary dns server was up for 500 days. Now I see in top that it's been up for only 3 days. No trace of reboot neither in messages nor in last log. Does top start over at 0 if a certain amount m> of up-days have been reached? Or how do I have to interprete that? m> thanks a lot m> Philipp http://uptime.netcraft.com/up/accuracy.html#cycle ... Additionally HP-UX, Linux, Solaris and recent releases of FreeBSD cycle back to zero after 497 days, exactly as if the machine had been rebooted at that precise point. Thus it is not possible to see a HP-UX, Linux or Solaris system with an uptime measurement above 497 days. -- Best regards, YANDEX mailto:student55@yandex.ru
participants (5)
-
mailinglists
-
Markus Gaugusch
-
Robert Davies
-
Roland Kuhn
-
YANDEX