John, It looks like Jared was correct, we had another instance of rancid running on another box a roughly the same time, I shifted the the cron job forward 30 minutes, and the diffs have stopped.
Kind Regards, Ben Cooper Support and Infrastructure Manager Hub Network Services Ltd Phone: 0117 9200045 Fax: 0127 5394292 For system or network support, please email supp...@hns.net On 19/03/2010 17:18, john heasley wrote: > Fri, Mar 19, 2010 at 10:40:20AM -0400, Jared Mauch: >> This typically happens if someone is viewing the startup-config (eg: show >> conf) as it is locked. > > afaict, reading nor writing locks the nvram fsys in such a way that > dir /all nvram:, the command rancid uses, fails. it seems to wait > as you'd expect the locking to work, though not in a fifo manner. > at least, i can't reproduce this on 12.2.18SXF16. > > if it did fail, i'd expect that it the cli would return an error, > which rancid may not recognize. in cases where the fsys is locked, > like for one being squeezed, the cli normally produces errors like > > Open device \S+ failed > Error opening \S+: > > which rancid does recognize. > > it may be that write memory removes all the nvram files, then writes > the new ones and you're just lucky. other ideas? > >> - Jared >> >> On Mar 19, 2010, at 7:58 AM, Ben Cooper wrote: >> >>> Hi, >>> >>> We use rancid to retrieve configs from our cisco kit, recently one of >>> our 6500s (s72033_rp-ADVENTERPRISEK9_WAN-M Version 12.2(33)SXH3) has >>> started reporting nvram content changes sporadically throughout the day, eg: >>> >>>> !Flash: nvram: Directory of nvram:/ >>>> !Flash: nvram: 1918 -rw- 26788 <no date> >>>> startup-config >>>> !Flash: nvram: 1919 ---- 24 <no date> >>>> private-config >>>> !Flash: nvram: 1920 -rw- 26788 <no date> >>>> underlying-config >>>> - !Flash: nvram: 1 ---- 4 <no date> >>>> rf_cold_starts >>>> - !Flash: nvram: 2 ---- 48 <no date> >>>> persistent-data >>>> - !Flash: nvram: 3 -rw- 4887 <no date> >>>> ifIndex-table >>>> !Flash: nvram: 1964024 bytes total (1929992 bytes free) >>> >>>> !Flash: nvram: Directory of nvram:/ >>>> - !Flash: nvram: No files in directory >>>> + !Flash: nvram: 1918 -rw- 26788 <no date> >>>> startup-config >>>> + !Flash: nvram: 1919 ---- 24 <no date> >>>> private-config >>>> + !Flash: nvram: 1920 -rw- 26788 <no date> >>>> underlying-config >>>> + !Flash: nvram: 1 ---- 4 <no date> >>>> rf_cold_starts >>>> + !Flash: nvram: 2 ---- 48 <no date> >>>> persistent-data >>>> + !Flash: nvram: 3 -rw- 4887 <no date> >>>> ifIndex-table >>>> !Flash: nvram: 1964024 bytes total (1929992 bytes free) >>> >>> Has anyone experienced this behaviour before? >>> >>> Thanks, >>> >>> Ben >>> _______________________________________________ >>> cisco-nsp mailing list cisco-nsp@puck.nether.net >>> https://puck.nether.net/mailman/listinfo/cisco-nsp >>> archive at http://puck.nether.net/pipermail/cisco-nsp/ >> >> >> _______________________________________________ >> cisco-nsp mailing list cisco-nsp@puck.nether.net >> https://puck.nether.net/mailman/listinfo/cisco-nsp >> archive at http://puck.nether.net/pipermail/cisco-nsp/ > > _______________________________________________ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/