expireover problems with ovdb?

Heath Kehoe heath.kehoe at intermec.com
Thu Sep 14 15:56:37 UTC 2000


>
>Morning all ...
>
>	I just switched over to using ovdb vs buffindexed, and it appears
>that expireover gets "stuck", and locks up the whole system (news system,
>not operating system) ...
>
>	A gcore of the expireover process reveals:
>
>(gdb) bt
>#0  0x280d7532 in __db_shalloc_free () from /usr/local/lib/libdb3.1.so
>#1  0x280f7c62 in __memp_mf_discard () from /usr/local/lib/libdb3.1.so
>#2  0x280f6860 in __memp_bhfree () from /usr/local/lib/libdb3.1.so
>#3  0x280f5dc9 in __memp_alloc () from /usr/local/lib/libdb3.1.so
>#4  0x280f6d83 in memp_fget () from /usr/local/lib/libdb3.1.so
>#5  0x280b5343 in __bam_c_next () from /usr/local/lib/libdb3.1.so
>#6  0x280b429c in __bam_c_get () from /usr/local/lib/libdb3.1.so
>#7  0x280ce529 in __db_c_get () from /usr/local/lib/libdb3.1.so
>#8  0x28081e83 in ovdb_expiregroup () at ovdb.c:987
>#9  0x2806b90e in OVexpiregroup () at ov.c:312
>#10 0x80490f6 in main (argc=3, argv=0xbfbff074) at expireover.c:167
>#11 0x8048ab9 in _start ()
>
>	If that helps ... a truss of the process just sits there, nothing
>coming up on the screen ...
>

Which version of BerkeleyDB are you using?  3.1.14 is known to
have bugs in its memp* stuff that can cause that.  Upgrade to
3.1.17 if you havn't already.

-heath




More information about the inn-workers mailing list