Apache/mod_ssl crashing on startup

Barry Roberts blr at robertsr.us
Wed Mar 18 13:29:19 MDT 2009


I have an internal-use-only web server that had several hundred days of 
uptime before a power change recently.  Since it has powered back up I 
can't get apache to start.

It's a 32-bit CentOS 5 machine.  If I remove the ssl.conf from 
/etc/httpd/conf.d, it starts up fine.  With it, it segfaults on startup.

httpd -t says the config files are syntactically correct.

I've yum updated (hoping I did something stupid on that machine in the 
last year) to see if that would fix it.  Nope.  Also rpm -V on httpd, 
mod_ssl, and openssl don't show anything unusual, just config files 
modified.

My ssl.conf is down to:
LoadModule ssl_module modules/mod_ssl.so

Listen 443

AddType application/x-x509-ca-cert .crt
AddType application/x-pkcs7-crl    .crl
SSLPassPhraseDialog  builtin
SSLSessionCache none


#SSLMutex default
SSLMutex none
SSLRandomSeed startup file:/dev/urandom  256
SSLRandomSeed connect builtin
SSLCryptoDevice builtin

<VirtualHost _default_:443>

SSLEngine on
SSLCertificateFile /etc/pki/tls/certs/localhost.crt
SSLCertificateKeyFile /etc/pki/tls/private/localhost.key

</VirtualHost>

strace httpd -k start shows the last system calls (I can send the full 
output) as either a whole bunch of:
time(NULL)                              = 1237404201

sometimes followed by some semaphore calls:
semget(IPC_PRIVATE, 1, IPC_CREAT|0600)  = 983052
semctl(983052, 0, IPC_64|SETVAL, 0xbff9d028) = 0
geteuid32()                             = 0
semctl(983052, 0, IPC_64|IPC_SET, 0xbff9cff8) = 0
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++

I've googled using those system calls as keywords and haven't found any 
similar problems posted.

Anybody have any idea what I can try next to get this server up and 
running again?  Short of re-installing.  That's SO Windows.

Thanks,
Barry





More information about the PLUG mailing list