Good day.
Openldap 2.4.18 with bdb backend used for pam_ldap and nss_ldap on freebsd7.2 amd64 Crashed every 10 - 30 minutes on serches. Load isnt heavy, about 10 req/sec in peak and 20req/min average.
Crash probability is higher when over program use the disk.
how to mo debug to resolve problem?
Some details:
#uname -a FreeBSD bazar 7.2-RELEASE FreeBSD 7.2-RELEASE #3: Tue Jun 9 16:46:20 MSD 2009 root@bazar:/usr/obj/usr/src/sys/GENERIC amd64
#/usr/local/libexec/slapd -V @(#) $OpenLDAP: slapd 2.4.18 (Oct 7 2009 13:29:25)
tuning in slapd.conf cachesize 150000 idlcachesize 450000 #checkpoint 64 10 threads 8 concurrency 5
tuning in /var/db/openldap-data/DB_CONFIG set_cachesize 0 62428800 0 set_lg_regionmax 1048576 set_lg_max 10485760 set_lg_bsize 2097152 set_lg_dir /var/db/openldap-data
crash debug whith truss /usr/local/libexec/slapd -d -1 _umtx_op(0x8022020b0,0x11,0x0,0x0,0x0,0x0) = 0 (0x0) gidNumber_umtx_op(0x8022020b0,0x11,0x0,0x0,0x0,0x7ffffda79308) = 10 (0xa) getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) sendto(3,"<167>Oct 7 11:59:33 slapd[81516"...,91,0x0,NULL,0x0) = 91 (0x5b) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7ffffebfba28) = 0 (0x0) ber_scanf fmt ({im}) ber: write(2,"ber_scanf fmt ({im}) ber:\n",26) = 0 (0x0) ber_dump: buf=0x81141b408 ptr=0x81141b408 end=0x81141b414 len=12 gettimeofday({1254902373.820741 },0x0) = 65 (0x41) 0000: 30 84 00 00 00 06 02 02 03 e8 04 00 0........... write(2," 0000: 30 84 00 00 00 06 02 02"...,79) = 0 (0x0) <= get_ctrls: n=1 rc=0 err="" write(2,"<= get_ctrls: n=1 rc=0 err=""\n",30) = 30 (0x1e) getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) sendto(3,"<167>Oct 7 11:59:33 slapd[81516"...,106,0x0,NULL,0x0) = 106 (0x6a) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7fffff3fcaa8) = 0 (0x0) gettimeofday({1254902373.821208 },0x0) = 0 (0x0) ==> limits_get: conn=92 op=1 self="cn=bind,ou=staff,dc=ulgsm,dc=ru" this="ou=staff,dc=ulgsm,dc=ru" getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) _umtx_op(0x8022020b0,0x11,0x0,0x0,0x0,0x7fffff3fc648) = 105 (0x69) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7ffffda79658) = 0 (0x0) gettimeofday({1254902373.821644 },0x0) = 34669068288 (0x812700000) getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) => bdb_dn2idl("ou=staff,dc=ulgsm,dc=ru") sendto(3,"<167>Oct 7 11:59:33 slapd[81516"...,45,0x0,NULL,0x0) = 41 (0x29) _umtx_op(0x8022020b0,0x11,0x0,0x0,0x0,0x7ffffda79208) = 45 (0x2d) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7ffffe3faaa8) = 0 (0x0)
write(2,"\n",1) = 0 (0x0) _umtx_op(0x8022020b0,0x11,0x0,0x0,0x0,0x7ffffe3fa6e8) = 0 (0x0) getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) sendto(3,"<167>Oct 7 11:59:33 slapd[81516"...,64,0x0,NULL,0x0) = 64 (0x40) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7ffffebfba28) = 0 (0x0) attrs:write(2," attrs:",10) = 0 (0x0) gettimeofday({1254902373.822603 },0x0) = 0 (0x0) getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) sendto(3,"<167>Oct 7 11:59:33 slapd[81516"...,45,0x0,NULL,0x0) = 45 (0x2d) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7ffffebfbaa8) = 0 (0x0) cngettimeofday({1254902373.822938 },0x0) = 3 (0x3) getpid(0x80180bf68,0x7fffffffed70,0x0,0x0,0xffff80080f10aaaa,0x0) = 81516 (0x13e6c) sendto(3,"<167>Oct 7 11:59:33 slapd[81516"...,75,0x0,NULL,0x0) = 75 (0x4b) _umtx_op(0x8022020b0,0x12,0x0,0x0,0x0,0x7ffffda795c8) = 0 (0x0) SIGNAL 11 (SIGSEGV)