<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8">
<META NAME="GENERATOR" CONTENT="GtkHTML/3.12.3">
</HEAD>
<BODY>
On Fri, 2007-03-09 at 04:20 -0500, Jeff Johnson wrote:<BR>
<BLOCKQUOTE TYPE=CITE>
<BR>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000">On Mar 8, 2007, at 7:19 PM, Richard Hendershot wrote:</FONT>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<BR>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000">On Thu, 2007-03-08 at 08:46 -0500, Jeff Johnson wrote: </FONT>
<BLOCKQUOTE TYPE=CITE>
<PRE>
<FONT COLOR="#000000">On Mar 8, 2007, at 7:23 AM, Neal Becker wrote:</FONT>
<FONT COLOR="#000000">> smart update hangs at ~50% loading cache. rm /var/lib/rpm/__*, then</FONT>
<FONT COLOR="#000000">> rpm --rebuilddb fixes it.</FONT>
<FONT COLOR="#000000">></FONT>
<FONT COLOR="#000000">You likley don''t need --rebuilddb.</FONT>
<FONT COLOR="#000000">This is likely the kernel-2.6.19 (and FC6 2.6.18 backport) mmap(2) </FONT>
<FONT COLOR="#000000">problem,</FONT>
<FONT COLOR="#000000">bugzilla is full of reports.</FONT>
</PRE>
</BLOCKQUOTE>
<BR>
<FONT COLOR="#000000">I *think* I had a similar problem. Removing /var/lib/smart/* seemed to correct the cache lockup. Make sense at all?? </FONT>
</BLOCKQUOTE>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<BR>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000">If doing</FONT>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000"> rm -f /var/lib/smart/*</FONT>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000">corrected, then the issue was different than rpmdbg stale locks with identical symptoms.</FONT>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<BR>
<BR>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000">Likelier (my guess) you did something like reboot, which does</FONT>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<FONT COLOR="#000000"> rm -f /var/lib/rpm/__db*</FONT>
</BLOCKQUOTE>
<BLOCKQUOTE TYPE=CITE>
<BR>
</BLOCKQUOTE>
<BR>
I didn't reboot. The issue that seemed similar to me was smart would consistently hang while updating its cache. I'd kill it and try rpm which would succeed (faster than looking into the docs for how/where it's semaphores lived - at least for me). My sites live in /etc/smart/channels which I strive to keep up-to-date so killing /var/lib/smart/config is not a problem. The GUI choices need to be reset but that's quick, and validates channel files.<BR>
<BR>
After removing the lib files smart was then able to process its cache and operate normally.<BR>
<BR>
I never had anything remotely approaching rpmdb corruption though so this may not be so closely related as I had though. YMMV<BR>
<TABLE CELLSPACING="0" CELLPADDING="0" WIDTH="100%">
<TR>
<TD>
<BR>
</TD>
</TR>
</TABLE>
</BODY>
</HTML>