[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: slapd crashing "randomly?"




On Apr 11, 2007, at 1:30 PM, Quanah Gibson-Mount wrote:

--On Wednesday, April 11, 2007 1:25 PM -0400 Daniel Henninger <daniel@ncsu.edu> wrote:

Been a while, but I finally caught a core dump. Of course, I'm not
entirely sure why there's so few useful symbols showing since I compiled
it with debugging symbols and didn't strip it. =/ Anyway, the
information I got from it is interesting:


(gdb) bt
# 0 0x000b4694 in ?? ()
# 1 0x000e175c in avl_delete ()
# 2 0x000b4c48 in bdb_idl_cache_put ()
# 3 0x000b5930 in bdb_idl_fetch_key ()
# 4 0x000b796c in bdb_key_read ()
# 5 0x000b30b0 in bdb_filter_candidates ()
# 6 0x000b3a28 in ?? ()
# 7 0x000b3a28 in ?? ()
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
(gdb)


Is it possible I just have a busted version of berkeley db?! What
version are you all using? (I guess it's Oracle DB now...) We are using
version 4.2.52. Built with --enable-compat185.

Hi Daniel,

What does "file slapd" say? In general, "make install" will strip the symbols from slapd even if you built it with debugging etc.

Aww crap! =(
/local/ldap/libexec/slapd: ELF 32-bit MSB executable, SPARC, version 1 (SYSV), dynamically linked (uses shared libs), stripped


Didn't think about that, thanks!

As for BDB 4.2.52, you must apply the patches from Oracle as well, otherwise it is known to corrupt.

Any newer ones suuggested over 4.2.52? I have no qualms at all with upgrading, just don't want to dive into a known busticated version. =)


Daniel


--Quanah


-- Quanah Gibson-Mount Senior Systems Software Developer ITS/Shared Application Services Stanford University GnuPG Public Key: http://www.stanford.edu/~quanah/pgp.html