[Date Prev] [Date Index] [Date Next] [Thread Prev] [Thread Index] [Thread Next]
Bryan Stansell bryan@conserver.com
Tue, 14 Feb 2006 10:19:48 -0800 (PST)
ah, this is just more of a misunderstanding. the timestamp is GMT-time (ok, maybe technically UTC time - but it's called gmtime(), so i'm callin' it GMT). the reason for it is that clocks can roll back for certain timezones and tracking localtime could (in theory) have logfiles "out of order" around time shifts as well as (very theoretical) clobber an older file as the clock rolls back. so, i thought it better to make it GMT-time and guarantee and always-increasing timestamp (assuming your clock isn't stepped back by ntpd or by hand or...). Bryan On Tue, Feb 14, 2006 at 12:58:25PM -0500, Mike Daigle wrote: > Another setup problem. Is this a bug? > > I have setup logfile rotation with logfilemax 3k; > I was looking for logs from the console og2. I see > # ls -al > total 166 > drwxr-xr-x 2 root sys 5632 Feb 14 10:43 . > drwxr-xr-x 31 root sys 512 Nov 17 13:40 .. > -rw-r--r-- 1 root other 552 Feb 14 10:45 celes.ccs > -rw-r--r-- 1 root other 70278 Feb 1 14:06 > celes.ccs-20060201-190639 > -rw-r--r-- 1 root other 1822 Feb 14 11:34 og2 > -rw-r--r-- 1 root other 3225 Feb 14 10:43 og2-20060214-154300 > # date > Tue Feb 14 12:53:47 EST 2006 > # > > Note there is a file called og2-20060214-154300 but the date shows we > have not yet reached the timestamp when that log says it was created. > Looking at the contents of this file it looks like the > timestamp/filename should be > og2-20060214-104300 > > Help?