Need help with core dump gdb output 2.1.3 server
Posted: Thu Oct 14, 2010 8:10 am
Hello,
Environment: OpenVPN 2.1.3 (non-standard TCP port) on RedHat 5.4el-64bit
Maximum number of users is about 10, mixture of PC and Mac clients.
I am having periodic crashes with core dumps enabled:
-rw------- 1 root root 2105344 Sep 22 14:31 core.13382
-rw------- 1 root root 3964928 Oct 8 16:25 core.17164
-rw------- 1 root root 2519040 Oct 14 14:24 core.17935
-rw------- 1 root root 4050944 Oct 4 15:04 core.6679
They all have gdb output something like:
# gdb /usr/local/sbin/openvpn core.17935
...< much stuff snipped here >...
Reading symbols from /lib64/libaudit.so.0...(no debugging symbols found)...done.
Loaded symbols for /lib64/libaudit.so.0
Core was generated by `/usr/local/sbin/openvpn --cd /etc/openvpn --daemon --config server.conf'.
Program terminated with signal 11, Segmentation fault.
#0 do_compress (in=<value optimized out>, in_len=1406, out=0x11a4e34 "", out_len=0x7fff7362ccc4, wrkmem=0x12314e0)
at lzo1x_c.ch:86
86 lzo1x_c.ch: No such file or directory.
in lzo1x_c.ch
(gdb)
All four of the above core dumps have the Seg. fault and the lzo1x_c.ch: No such file... message.
I'm not sure if this is an LZO problem or what.
server.conf is below.
I do not see any LZO errors in the openvpn.log file.
The OS is 64 bit and so are the lzo libraries:
# file /usr/lib64/liblzo.so.1.0.0
/usr/lib64/liblzo.so.1.0.0: ELF 64-bit LSB shared object, AMD x86-64, version 1 (SYSV), not stripped
# file /usr/lib64/liblzo2.so.2.0.0
/usr/lib64/liblzo2.so.2.0.0: ELF 64-bit LSB shared object, AMD x86-64, version 1 (SYSV), not stripped
I have a script which checks connectivity and before and after restarting OpenVPN does a ps ax | grep openvpn
The output from the last crash was:
****** before restart *******
11755 ? R 0:00 grep openvpn
17930 ? S 0:00 /usr/local/sbin/openvpn --cd /etc/openvpn --daemon
****** after restart *******
11763 ? S 0:00 /usr/local/sbin/openvpn --cd /etc/openvpn --daemon --config server.conf
11769 ? Ss 0:00 /usr/local/sbin/openvpn --cd /etc/openvpn --daemon --config server.conf
11771 ? S 0:00 grep openvpn
Does anybody have any suggestions on how I should proceed with troubleshooting this problem? What is most perplexing is that it only happens every four to ten days or so.
Thanks.
Randy
server.conf:
# cat server.conf
dev tun
proto tcp
port XXXX (a non-standard port)
keepalive 10 120
server 10.x.x.0 255.255.255.0
duplicate-cn
ca ca.crt
cert server.crt
key server.key
dh dh1024.pem
plugin /etc/openvpn/openvpn-auth-pam.so login
client-cert-not-required
username-as-common-name
script-security 2
comp-lzo
verb 4
persist-key
persist-tun
status openvpn-status.log
log-append /var/log/openvpn.log
Environment: OpenVPN 2.1.3 (non-standard TCP port) on RedHat 5.4el-64bit
Maximum number of users is about 10, mixture of PC and Mac clients.
I am having periodic crashes with core dumps enabled:
-rw------- 1 root root 2105344 Sep 22 14:31 core.13382
-rw------- 1 root root 3964928 Oct 8 16:25 core.17164
-rw------- 1 root root 2519040 Oct 14 14:24 core.17935
-rw------- 1 root root 4050944 Oct 4 15:04 core.6679
They all have gdb output something like:
# gdb /usr/local/sbin/openvpn core.17935
...< much stuff snipped here >...
Reading symbols from /lib64/libaudit.so.0...(no debugging symbols found)...done.
Loaded symbols for /lib64/libaudit.so.0
Core was generated by `/usr/local/sbin/openvpn --cd /etc/openvpn --daemon --config server.conf'.
Program terminated with signal 11, Segmentation fault.
#0 do_compress (in=<value optimized out>, in_len=1406, out=0x11a4e34 "", out_len=0x7fff7362ccc4, wrkmem=0x12314e0)
at lzo1x_c.ch:86
86 lzo1x_c.ch: No such file or directory.
in lzo1x_c.ch
(gdb)
All four of the above core dumps have the Seg. fault and the lzo1x_c.ch: No such file... message.
I'm not sure if this is an LZO problem or what.
server.conf is below.
I do not see any LZO errors in the openvpn.log file.
The OS is 64 bit and so are the lzo libraries:
# file /usr/lib64/liblzo.so.1.0.0
/usr/lib64/liblzo.so.1.0.0: ELF 64-bit LSB shared object, AMD x86-64, version 1 (SYSV), not stripped
# file /usr/lib64/liblzo2.so.2.0.0
/usr/lib64/liblzo2.so.2.0.0: ELF 64-bit LSB shared object, AMD x86-64, version 1 (SYSV), not stripped
I have a script which checks connectivity and before and after restarting OpenVPN does a ps ax | grep openvpn
The output from the last crash was:
****** before restart *******
11755 ? R 0:00 grep openvpn
17930 ? S 0:00 /usr/local/sbin/openvpn --cd /etc/openvpn --daemon
****** after restart *******
11763 ? S 0:00 /usr/local/sbin/openvpn --cd /etc/openvpn --daemon --config server.conf
11769 ? Ss 0:00 /usr/local/sbin/openvpn --cd /etc/openvpn --daemon --config server.conf
11771 ? S 0:00 grep openvpn
Does anybody have any suggestions on how I should proceed with troubleshooting this problem? What is most perplexing is that it only happens every four to ten days or so.
Thanks.
Randy
server.conf:
# cat server.conf
dev tun
proto tcp
port XXXX (a non-standard port)
keepalive 10 120
server 10.x.x.0 255.255.255.0
duplicate-cn
ca ca.crt
cert server.crt
key server.key
dh dh1024.pem
plugin /etc/openvpn/openvpn-auth-pam.so login
client-cert-not-required
username-as-common-name
script-security 2
comp-lzo
verb 4
persist-key
persist-tun
status openvpn-status.log
log-append /var/log/openvpn.log