Problems with inn 2.2.3

Mark Hittinger bugs at pu.net
Mon Apr 29 22:30:49 UTC 2002




| From: ptomblin at xcski.com (Paul Tomblin)
| I rebuild the history file, and it's setting down to a much more
| reasonable 26Mb.  Running news.daily got rid of most of the duplicates.
| I've still got a bunch of those history items like 
| [5D5978E11CE8882699819BFCA7D57129]      1020115686~-~1020115686
| [33C9E1199C2DCAC21BBA28F0077B2694]      1020115724~-~1020115724
| [7603278280298CBE026E6B1DD47C547D]      1020115851~-~1020115851
| mixed in amongst the more normal
| <aakd2k$ctg$1 at vulcan.alphanet.ch>	1020116109~-~1020114836	rec.aviation.student/92238
| <aake4k$ctg$2 at vulcan.alphanet.ch>	1020116109~-~1020115924	rec.aviation.student/92239
| <aakea7$bks7h$1 at ID-121350.news.dfncis.de>	1020116111~-~1020115254	alt.lang.asm/2414

Aaaaaaaaahhh!!!  What a scary looking history file! :-)

Those "normal" records do not look normal for inn 2.x to me.  I've not run 2.x
with traditional spool so I am not 100% sure about this but my belief is that
2.x history file records never have the message id in field #1.  Its always
the hash so that things like the history cache can work etc.

Is it possible you have some 1.x binaries putting entries into the history
file somehow?

Later

Mark Hittinger
bugs at pu.net


More information about the inn-bugs mailing list