View previous topic :: View next topic |
Author |
Message |
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Mon Nov 26, 2012 10:10 am Post subject: apache error |
|
|
Hi all,
Does any one have a clue about solving this error:
Quote: |
[Mon Nov 26 09:27:52 2012] [notice] Apache/2.2.16 (Unix) mod_ssl/2.2.16 OpenSSL/0.9.8n DAV/2 PHP/5.3.3 configured -- resuming normal operations
zend_mm_heap corrupted
[Mon Nov 26 09:41:06 2012] [notice] child pid 19448 exit signal Segmentation fault (11)
[Mon Nov 26 09:41:35 2012] [notice] child pid 19423 exit signal Segmentation fault (11)
[Mon Nov 26 09:44:19 2012] [notice] child pid 19402 exit signal Segmentation fault (11)
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
[Mon Nov 26 09:48:36 2012] [notice] child pid 19453 exit signal Segmentation fault (11)
zend_mm_heap corrupted
[Mon Nov 26 09:48:46 2012] [notice] child pid 19638 exit signal Segmentation fault (11)
[Mon Nov 26 09:48:50 2012] [notice] child pid 19627 exit signal Segmentation fault (11)
zend_mm_heap corrupted
[Mon Nov 26 09:48:53 2012] [notice] child pid 19706 exit signal Segmentation fault (11)
[Mon Nov 26 09:50:43 2012] [notice] child pid 19721 exit signal Segmentation fault (11)
[Mon Nov 26 09:53:21 2012] [notice] child pid 19725 exit signal Segmentation fault (11)
[Mon Nov 26 09:55:40 2012] [notice] child pid 19718 exit signal Segmentation fault (11)
|
I've tried the solution here by increasing output_buffering to 8192(double as default) but unefective.
PHP details:
Quote: |
php -v
PHP 5.3.3 (cli) (built: Aug 6 2010 21:17:10)
Copyright (c) 1997-2010 The PHP Group
Zend Engine v2.3.0, Copyright (c) 1998-2010 Zend Technologies
|
|
|
Back to top |
|
gerasimos_h Site Admin
Joined: 09 Aug 2007 Posts: 1757 Location: Greece
|
Posted: Mon Nov 26, 2012 1:50 pm Post subject: |
|
|
Have you try other suggestions at that post?
Also by default session save path is at /var/lib/php
take a look at there (permissions, quota e.t.c)
Can you upgrade php?
What is your SMS version?
I can reproduce segfault with 5.3.19 (& 5.4.x) but not corruption.
Another solution is to use memcache
http://slackbuilds.org/repository/14.0/network/memcached/
SMS include the php extension by default.
I'm also thinking adding memcache in SMS.
gerasimos_h
_________________ Superb! Mini Server Project Manager
http://sms.it-ccs.com |
|
Back to top |
|
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Mon Nov 26, 2012 2:42 pm Post subject: |
|
|
Hi,
Thanks for your answer.
I only increased output_buffering to 8192 in php.ini with no efect.
SMS version is 1.5.1. installed since 06.sep.2011.
/var/lib/php seems ok as it is untouched and contain already some sess_* files.
I can give a try to memcache since it seem to be helpfull.
There is something else I found in the error_log, you can find in the attachment.
Also I noticed that the first occurence of
Quote: | file: could not find any magic files! |
was:
Quote: |
[Sat Nov 05 10:15:28 2011] [error] [client 79.117.113.127] Invalid URI in request HTTP/1.1 200 OK
[Sat Nov 05 10:15:28 2011] [error] [client 79.117.113.127] Invalid URI in request HTTP/1.1 200 OK
[Sat Nov 05 10:34:07 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 11:04:06 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 11:34:05 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 12:04:04 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 12:15:47 2011] [error] [client 79.117.114.254] request failed: error reading the headers
[Sat Nov 05 12:15:47 2011] [error] [client 79.117.114.254] request failed: error reading the headers
[Sat Nov 05 12:34:04 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 13:04:06 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 13:34:05 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 14:04:05 2011] [error] [client 109.166.139.252] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
[Sat Nov 05 14:04:56 2011] [error] [client 87.230.74.47] File does not exist: /var/www/din.aspx
[Sat Nov 05 14:22:07 2011] [error] [client 109.166.136.251] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
file: could not find any magic files!
file: could not find any magic files!
file: could not find any magic files!
file: could not find any magic files!
file: could not find any magic files!
file: could not find any magic files!
[Sat Nov 05 18:36:33 2011] [error] [client 61.150.114.76] Directory index forbidden by Options directive: /var/www/
file: could not find any magic files!
|
and te last:
Quote: |
[Thu Nov 01 23:31:21 2012] [notice] child pid 11543 exit signal Segmentation fault (11)
zend_mm_heap corrupted
file: could not find any magic files!
|
First occurence of
Quote: |
zend_mm_heap corrupted
|
was
Quote: |
[Tue Oct 30 04:40:07 2012] [notice] Apache/2.2.16 (Unix) mod_ssl/2.2.16 OpenSSL/0.9.8n DAV/2 PHP/5.3.3 configured -- resuming normal operations
zend_mm_heap corrupted
|
Description: |
|
Download |
Filename: |
error_log_excerpt.txt |
Filesize: |
63.91 KB |
Downloaded: |
1547 Time(s) |
|
|
Back to top |
|
gerasimos_h Site Admin
Joined: 09 Aug 2007 Posts: 1757 Location: Greece
|
Posted: Mon Nov 26, 2012 5:09 pm Post subject: |
|
|
What exactly is running and requires 'file' command or posdrop.
Did you configure web app to look for magic files, or if require a mine that isn't listed?
Magic files are in /etc/file, but you can add symlinks if web apps look elsewhere like
Code: | /etc/magic -> file/magic/
/etc/magic.mime -> /etc/file/magic.mime* |
Also" zend_mm_heap corrupted" could occurred from a php app.
gerasimos_h
_________________ Superb! Mini Server Project Manager
http://sms.it-ccs.com |
|
Back to top |
|
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Mon Nov 26, 2012 6:35 pm Post subject: |
|
|
Not something I know.
It looks stange to me and this is why I mentioned the excerpt form error_log.
I have Postfix running, fail2ban. mailscanner(clamav&sa) and vsftpd.
Can't imagine who can require file.
I am also hosting several websites but html mostly.
|
|
Back to top |
|
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Mon Nov 26, 2012 7:27 pm Post subject: |
|
|
now I'm getting strange messages in console:
Quote: |
root@aiolos:/var/log/httpd# tail -f /var/log/httpd/error_log
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
zend_mm_heap corrupted
Message from syslogd@aiolos at Mon Nov 26 19:18:18 2012 ...
aiolos kernel: nf_ct_snmp: dropping packetIN= OUT=eth0 SRC=83.103.186.217 DST=92.239.15.55 LEN=58 TOS=0x00 PREC=0x00 TTL=127 ID=1070 PROTO=UDP SPT=46511 DPT=161 LEN=38
Message from syslogd@aiolos at Mon Nov 26 19:18:21 2012 ...
aiolos kernel: nf_ct_snmp: dropping packetIN= OUT=eth0 SRC=83.103.186.217 DST=92.239.15.55 LEN=58 TOS=0x00 PREC=0x00 TTL=127 ID=1380 PROTO=UDP SPT=46511 DPT=161 LEN=38
Message from syslogd@aiolos at Mon Nov 26 19:18:28 2012 ...
aiolos kernel: nf_ct_snmp: dropping packetIN= OUT=eth0 SRC=83.103.186.217 DST=92.239.15.55 LEN=58 TOS=0x00 PREC=0x00 TTL=127 ID=2033 PROTO=UDP SPT=46511 DPT=161 LEN=38
zend_mm_heap corrupted
Message from syslogd@aiolos at Mon Nov 26 19:19:50 2012 ...
aiolos kernel: nf_ct_snmp: dropping packetIN= OUT=eth0 SRC=83.103.186.217 DST=92.239.15.55 LEN=58 TOS=0x00 PREC=0x00 TTL=127 ID=12941 PROTO=UDP SPT=46511 DPT=161 LEN=38
Message from syslogd@aiolos at Mon Nov 26 19:19:53 2012 ...
aiolos kernel: nf_ct_snmp: dropping packetIN= OUT=eth0 SRC=83.103.186.217 DST=92.239.15.55 LEN=58 TOS=0x00 PREC=0x00 TTL=127 ID=13579 PROTO=UDP SPT=46511 DPT=161 LEN=38
Message from syslogd@aiolos at Mon Nov 26 19:20:00 2012 ...
aiolos kernel: nf_ct_snmp: dropping packetIN= OUT=eth0 SRC=83.103.186.217 DST=92.239.15.55 LEN=58 TOS=0x00 PREC=0x00 TTL=127 ID=14564 PROTO=UDP SPT=46511 DPT=161 LEN=38
zend_mm_heap corrupted
|
|
|
Back to top |
|
gerasimos_h Site Admin
Joined: 09 Aug 2007 Posts: 1757 Location: Greece
|
Posted: Mon Nov 26, 2012 8:07 pm Post subject: |
|
|
That's not from httpd/error_log it's from kernel, an emergency message
Check fail2ban.log if it banned the specified IP.
Have you check scripts that running from your hosters.
gerasimos_h
_________________ Superb! Mini Server Project Manager
http://sms.it-ccs.com |
|
Back to top |
|
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Mon Nov 26, 2012 8:42 pm Post subject: |
|
|
You're right, but the message comes from my rc.firewall not from fail2ban.
I have stopped the httpd service but "zend_mm_heap corrupted" still appear in /var/log/httpd/error_log.
|
|
Back to top |
|
gerasimos_h Site Admin
Joined: 09 Aug 2007 Posts: 1757 Location: Greece
|
Posted: Mon Nov 26, 2012 8:53 pm Post subject: |
|
|
Are all httpd process stopped?
If they are then maybe it's an emergency message too.
Try to comment *.emerg in /etc/syslog.conf and restart syslog too see.
gerasimos_h
_________________ Superb! Mini Server Project Manager
http://sms.it-ccs.com |
|
Back to top |
|
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Mon Nov 26, 2012 9:10 pm Post subject: |
|
|
Yes all httpd instances are off.
I did as you said with syslog, but I don't think it's emergency message since it appears inside /var/log/httpd/error_log.
The other message is written in /var/log/syslog and was displayed on the screen because I was monitoring /var/log/httpd/error_log with "tail -f"
I'll keep httpd off, now that I checked that all instances are stopped, to see if the message still appears. First time when I stopped httpd, I didn't check that all instances are down.
If nothing shows up till tomorrow, I'll start apache but rise online the websites one by one to see if any triggers the zend error.
|
|
Back to top |
|
keopp Senior Member
Joined: 08 Nov 2008 Posts: 166 Location: Romania
|
Posted: Tue Nov 27, 2012 6:20 pm Post subject: |
|
|
Problem solved!
There was one of the websites hosted.
The relevant error I found in the website-error_log was:
Quote: |
PHP Fatal error: Cannot redeclare zecimal() (previously declared in /var/www...
|
After I solved the multiple redundant declarations of that function, everything retured to normal, no zend errors (I'll still watch for magic file errors... ).
Thanks again for support.
|
|
Back to top |
|
|