You need 64-bit counter and there is none for uptime. Since once every 500 days is rare people just deal with such extra alert. Most people for one reason or another reboot their servers more often though.
If you want to write a work-around it would be one that uses previous run uptime data and if its near wrap time ignores the situation. On Thu, Oct 23, 2014 at 12:06 AM, Florian Lohoff <[email protected]> wrote: > > Hi, > i am currently monitoring sysuptime via check-snmp > > command_line $USER1$/check_snmp -H $HOSTADDRESS$ -C > $_HOSTCOMMUNITY$ -P 2c -o .1.3.6.1.2.1.1.3.0 -c 90000: > > The 90000 is basically that i want the first 15 Minutes after boot the > alarm > to be critical. > > Now obviously with a 32 bit counter this counter wraps at least every 497 > days. > > flo@p2:~$ echo $[ 2**32 / 100 / 86400 ] > 497 > > This causes the above statement to actually show a reboot/sysuptime ALARM. > > I dont see any way around this as the wrap depends on the last value found > via check_snmp so it would need some kind of persistence of data. > > Is there a better sysuptime check? > > Flo > -- > Florian Lohoff [email protected] > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > > iQIVAwUBVEio7JDdQSDLCfIvAQpycQ//ayLcurUViD5vBKpZ/6zVebm7y7pxtXL/ > SNW6pac9ls0y8gsJjSolHrVobHR+Xtdz/gl2pA86wI4+l1txJlAirpocuw8QHtam > Hg39ndTbryvDkxekUdxJi0Alue21kKqMHGnRj8B5Utpd8YLqQ6WhGQVHZR4Qpsiq > m49BEmCJyvQsQZPfOm4wgeAL1uqzXKIbxT1fiL4lghybFyKd8DE5QdzBGnEjJUSM > CEKMC/10ItSUd7CLaV6lfXBG+I1RbWJuyHPhpm8tbzvQtftnrk2PDG9cdjaMtSu1 > MVnnh46gMuiDzWK6CQPJiQEAeIdHGb1r2roPnUEjJT3vQvZLA9yIEFuwyVzAIA6V > aX5QZLhMoB0JVmU2sUMWqrZXBnrI7tLuHThdEQkOIvSfTbwNIAQsTwKSAwiLjjPC > BvnBrNNHu17S3n3KCMEzB1JZRT769ZpCkrheSGVjCj5eDcVf/dD0ENJndUWyZFOp > tXnB/7LkNlO4tO8ZrCStWSOBbXkohdUEVLNkvMRDniNWHsOgr36EdKLifMml11Mj > C7e8KsKUOPU9clydynbSosVRu62NR4ythwV1l2Tzgkq/SFI1QB704iA/Wc2rKzAz > ifwfi4NFH5RanWHWzgQSkWruEMs8U1dRbUmqPV5hiW/M4nHNt+4yLBlLHqtiJ9UQ > 0fhGGhnz9EQ= > =9UBV > -----END PGP SIGNATURE----- > >
