I have a VMware virtual machine with CentOS 6 32bit updated to the lates patches
Yesterday Apache started to give this error:
[Tue May 22 09:46:07 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:08 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket)
Any suggestion about this?
Ciao, luigi
On 5/22/2012 3:49 AM, Luigi Rosa wrote:
(38)Function not implemented
https://www.google.com/search?q=+%2838%29Function+not+implemented+apache
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Bob Hoffman said the following on 22/05/12 15:01:
(38)Function not implemented
https://www.google.com/search?q=+%2838%29Function+not+implemented+apache
Googled for two days before writing on this list.
Everything I found was about Gentoo compilation errors, old Suse issues and apr installation issues
Needless to say, I already reinstalled Apache and apr and disabled many modules.
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
Things should be made as simple as possible, but no simpler. --Albert Einstein
On 5/22/2012 3:49 AM, Luigi Rosa wrote:
I have a VMware virtual machine with CentOS 6 32bit updated to the lates patches
Yesterday Apache started to give this error:
[Tue May 22 09:46:07 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:08 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket)
Any suggestion about this?
Ciao, luigi
my guess is this is a non base package apache install or you changed something in the settings. you updated the kernel a few days ago and now an issue with file locking. that is what I got from the threads I read... I don't know about your set up though.
keep plugging away, whatever it is should be simple to fix once you find it.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Bob Hoffman said the following on 22/05/12 15:03:
On 5/22/2012 3:49 AM, Luigi Rosa wrote:
I have a VMware virtual machine with CentOS 6 32bit updated to the lates patches
Yesterday Apache started to give this error:
[Tue May 22 09:46:07 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket)
my guess is this is a non base package apache install or you changed something in the settings. you updated the kernel a few days ago and now an issue with file locking. that is what I got from the threads I read... I don't know about your set up though.
keep plugging away, whatever it is should be simple to fix once you find it.
Removed all php, apr and httpd packages
Removed /etc/httpd directory
Removed /var/www directory
Reinstalled with
yum install httpd php php-pear php-cli php-intl php-mysql php-imap php-mbstring php-xml php-pdo php-gd php-common
Started apache (no config changes)
Still getting the same error.
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
Bar Chart: A list of places to go when it's Miller time.
On Tue, May 22, 2012 at 9:52 AM, Luigi Rosa lists@luigirosa.com wrote:
Reinstalled with
yum install httpd php php-pear php-cli php-intl php-mysql php-imap php-mbstring php-xml php-pdo php-gd php-common
Started apache (no config changes)
Still getting the same error.
Do you have the stock apr-1.3.9-3.el6_1.2.x86_64 and apr-util-1.3.9-3.el6_0.1.x86_64 packages?
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Les Mikesell said the following on 22/05/12 17:22:
Do you have the stock apr-1.3.9-3.el6_1.2.x86_64 and apr-util-1.3.9-3.el6_0.1.x86_64 packages?
Yes, the x86 version because the server has an x86 CPU:
# rpm -q --all | grep apr apr-1.3.9-3.el6_1.2.i686 apr-util-ldap-1.3.9-3.el6_0.1.i686 apr-util-1.3.9-3.el6_0.1.i686
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
If a religion is defined to be a system of ideas that contains unprovable statements, then Goedel taught us that mathematics is not only a religion, it is the only religion that can prove itself to be one. --John Barrow
Luigi Rosa wrote:
Bob Hoffman said the following on 22/05/12 15:03:
On 5/22/2012 3:49 AM, Luigi Rosa wrote:
I have a VMware virtual machine with CentOS 6 32bit updated to the lates patches
Yesterday Apache started to give this error:
[Tue May 22 09:46:07 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket)
my guess is this is a non base package apache install or you changed something in the settings. you updated the kernel a few days ago and now an issue with file locking. that is what I got from the threads I
read... I
don't know about your set up though.
<snip>
Removed all php, apr and httpd packages
Removed /etc/httpd directory
Removed /var/www directory
Reinstalled with
yum install httpd php php-pear php-cli php-intl php-mysql php-imap php-mbstring php-xml php-pdo php-gd php-common
Started apache (no config changes)
Still getting the same error.
You might try reinstalling apr and apr-util, which may have the functions you need.
I see there's also an libapreq2, but I'd install the two I just mentioned first, then restart apache.
mark mark
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
m.roth@5-cent.us said the following on 22/05/12 17:23:
You might try reinstalling apr and apr-util, which may have the functions you need.
That's what I did when I removed Apache and php.
The yum install httpd installed apr for dependancy.
I see there's also an libapreq2, but I'd install the two I just mentioned first, then restart apache.
No libapreq2 file present
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
I predict that today will be remembered until tomorrow!
Luigi Rosa wrote:
m.roth@5-cent.us said the following on 22/05/12 17:23:
You might try reinstalling apr and apr-util, which may have the functions you need.
That's what I did when I removed Apache and php.
The yum install httpd installed apr for dependancy.
Ok. And yeah, I saw Les and I made the same suggestion within a minute of each other.
I see there's also an libapreq2, but I'd install the two I just mentioned first, then restart apache.
No libapreq2 file present
I don't know if it's needed for that. Sorry, I haven't run into this, and I'd have to research it.
mark
On 5/22/2012 11:35 AM, Luigi Rosa wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
m.roth@5-cent.us said the following on 22/05/12 17:23:
You might try reinstalling apr and apr-util, which may have the functions you need.
That's what I did when I removed Apache and php.
The yum install httpd installed apr for dependancy.
I see there's also an libapreq2, but I'd install the two I just mentioned first, then restart apache.
No libapreq2 file present
Ciao, luigi
/ +--[Luigi Rosa]-- \
I predict that today will be remembered until tomorrow! -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAk+7si0ACgkQ3kWu7Tfl6ZSaigCeOLVnJZg4yTGj96I5NkjaiGSo fWoAnjy1WEhxqqlLADXqa6QDVCn4QO8C =oGD9 -----END PGP SIGNATURE----- _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
yea , but uninstall makes a rpmsave that likes to add the previous changes (or least it seems to in some cases). I usually look for rpmsave, delete them then reinstall just in case.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
bob said the following on 22/05/12 17:48:
yea , but uninstall makes a rpmsave that likes to add the previous changes (or least it seems to in some cases). I usually look for rpmsave, delete them then reinstall just in case.
deleted /etc/httpd and /var/www after rpm -e to avoid this issue
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
Isn't it strange, G'Kar? When we first met, I had no power... and all the choices I could ever want. And now I have all the power I could ever want... and no choices at all. No choice... at all. --Londo, "The fall of Centauri Prime"
On Tue, 2012-05-22 at 18:02 +0200, Luigi Rosa wrote:
$ strings /usr/lib/libapr-1.so.0|grep apr_socket_accept apr_socket_accept
yum whatprovides /usr/lib/libapr-1.so.0 A link that points to? apr-..........el6
# Do not do a gracefull restart here! service httpd stop service hpptd start
ps aux|grep httpd # you want the parent proccess or pstree -p |grep httpd
strace -f -o httpd.txt -p 18289 '-p = parent_id_here'
Then kill strace after you see the error in the logs.
grep apr_socket_accept httpd.txt
I don't think you need to post the full strace output to the list, just the needed bits. It can get rather large. See man strace if you need to have strace do individual child process debugging in separate files.
That fails install debug packages and use gdb. You'll find the answer...
John
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
John Stanley said the following on 22/05/12 19:08:
$ strings /usr/lib/libapr-1.so.0|grep apr_socket_accept apr_socket_accept
Ok, the strings command gives the same result.
yum whatprovides /usr/lib/libapr-1.so.0 A link that points to? apr-..........el6
apr-1.3.9-3.el6_1.2.i686 : Apache Portable Runtime library Repo : base Matched from: Filename : /usr/lib/libapr-1.so.0
apr-1.3.9-3.el6_1.2.i686 : Apache Portable Runtime library Repo : installed Matched from: Other : Provides-match: /usr/lib/libapr-1.so.0
# Do not do a gracefull restart here! service httpd stop service hpptd start
Done.
ps aux|grep httpd # you want the parent proccess or pstree -p |grep httpd
strace -f -o httpd.txt -p 18289 '-p = parent_id_here'
Then kill strace after you see the error in the logs.
grep apr_socket_accept httpd.txt
got a lot of Process 32565 attached Process 32566 attached Process 32567 attached Process 32568 attached Process 32561 detached
grep returns an empty set, there is no 'apr' in the entire httpd.txt
Some (I suppose) relevant parts:
32373 select(0, NULL, NULL, NULL, {0, 31857}) = 0 (Timeout) 32373 clone(child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0xb78797b8) = 32434 32373 clone(child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0xb78797b8) = 32435 32434 rt_sigaction(SIGHUP, {0x5d4f00, [], SA_INTERRUPT}, {0x5d3d40, [HUP USR1], 0}, 8) = 0 32434 rt_sigaction(SIGTERM, {0x5d4f00, [], SA_INTERRUPT}, {0x5d3d00, [], 0}, 8) = 0 32434 rt_sigaction(SIGUSR1, {0x5d45a0, [], SA_INTERRUPT}, {0x5d3d40, [HUP USR1], 0}, 8) = 0 32434 geteuid32( <unfinished ...> 32373 waitpid(-1, <unfinished ...> 32434 <... geteuid32 resumed> ) = 0 32434 setgid32(48) = 0 32434 open("/proc/sys/kernel/ngroups_max", O_RDONLY <unfinished ...> 32373 <... waitpid resumed> [{WIFEXITED(s) && WEXITSTATUS(s) == 1}], WNOHANG|WSTOPPED) = 32432 32434 <... open resumed> ) = 8 32373 waitpid(-1, <unfinished ...> 32435 rt_sigaction(SIGHUP, {0x5d4f00, [], SA_INTERRUPT}, <unfinished ...> 32434 read(8, <unfinished ...> 32373 <... waitpid resumed> 0xbfa8ce3c, WNOHANG|WSTOPPED) = 0 32434 <... read resumed> "65536\n", 31) = 6 32373 select(0, NULL, NULL, NULL, {1, 0} <unfinished ...> 32435 <... rt_sigaction resumed> {0x5d3d40, [HUP USR1], 0}, 8) = 0 32434 close(8 <unfinished ...> 32435 rt_sigaction(SIGTERM, {0x5d4f00, [], SA_INTERRUPT}, <unfinished ...> 32434 <... close resumed> ) = 0 32435 <... rt_sigaction resumed> {0x5d3d00, [], 0}, 8) = 0 32434 open("/etc/group", O_RDONLY|O_CLOEXEC <unfinished ...> 32435 rt_sigaction(SIGUSR1, {0x5d45a0, [], SA_INTERRUPT}, <unfinished ...> 32434 <... open resumed> ) = 8 32435 <... rt_sigaction resumed> {0x5d3d40, [HUP USR1], 0}, 8) = 0 32434 fstat64(8, <unfinished ...> 32435 geteuid32( <unfinished ...> 32434 <... fstat64 resumed> {st_mode=S_IFREG|0644, st_size=981, ...}) = 0 32435 <... geteuid32 resumed> ) = 0 32434 mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0 <unfinished ...> 32435 setgid32(48 <unfinished ...> 32434 <... mmap2 resumed> ) = 0xb788d000 32435 <... setgid32 resumed> ) = 0 32434 _llseek(8, 0, <unfinished ...> 32435 open("/proc/sys/kernel/ngroups_max", O_RDONLY <unfinished ...> 32434 <... _llseek resumed> [0], SEEK_CUR) = 0 32435 <... open resumed> ) = 8 32434 read(8, <unfinished ...> 32435 read(8, <unfinished ...> 32434 <... read resumed> "root:x:0:root\nbin:x:1:root,bin,d"..., 4096) = 981 32435 <... read resumed> "65536\n", 31) = 6 32434 read(8, <unfinished ...> 32435 close(8 <unfinished ...> 32434 <... read resumed> "", 4096) = 0 32435 <... close resumed> ) = 0 32434 close(8 <unfinished ...> 32435 open("/etc/group", O_RDONLY|O_CLOEXEC <unfinished ...> 32434 <... close resumed> ) = 0 32435 <... open resumed> ) = 8 32434 munmap(0xb788d000, 4096 <unfinished ...> 32435 fstat64(8, <unfinished ...> 32434 <... munmap resumed> ) = 0
32435 <... read resumed> "root:x:0:root\nbin:x:1:root,bin,d"..., 4096) = 981 32434 epoll_create1(O_CLOEXEC <unfinished ...> 32435 read(8, <unfinished ...> 32434 <... epoll_create1 resumed> ) = 8 32435 <... read resumed> "", 4096) = 0 32434 epoll_ctl(8, EPOLL_CTL_ADD, 3, {EPOLLIN, {u32=3037114568, u64=3037114568}} <unfinished ...> 32435 close(8 <unfinished ...> 32434 <... epoll_ctl resumed> ) = 0 32435 <... close resumed> ) = 0 32434 accept4(3, <unfinished ...> 32435 munmap(0xb788d000, 4096 <unfinished ...> 32434 <... accept4 resumed> 0xbfa8ccfc, [128], SOCK_CLOEXEC) = -1 EINVAL (Invalid argument) 32435 <... munmap resumed> ) = 0 32434 accept4(-1, <unfinished ...> 32435 setgroups32(1, [48] <unfinished ...> 32434 <... accept4 resumed> 0xbfa8ccfc, [128], 0) = -1 EINVAL (Invalid argument) 32435 <... setgroups32 resumed> ) = 0 32434 gettimeofday( <unfinished ...> 32435 geteuid32( <unfinished ...> 32434 <... gettimeofday resumed> {1337707092, 778882}, NULL) = 0 32435 <... geteuid32 resumed> ) = 0 32434 write(2, "[Tue May 22 19:18:12 2012] [erro"..., 100 <unfinished ...> 32435 setuid32(48 <unfinished ...> 32434 <... write resumed> ) = 100 32435 <... setuid32 resumed> ) = 0 32434 close(8 <unfinished ...> 32435 epoll_create1(O_CLOEXEC <unfinished ...> 32434 <... close resumed> ) = 0 32435 <... epoll_create1 resumed> ) = 8 32434 munmap(0xcc3000, 84236 <unfinished ...> 32435 epoll_ctl(8, EPOLL_CTL_ADD, 3, {EPOLLIN, {u32=3037114568, u64=3037114568}}) = 0 32434 <... munmap resumed> ) = 0 32435 accept4(3, 0xbfa8ccfc, [128], SOCK_CLOEXEC) = -1 EINVAL (Invalid argument) 32434 munmap(0x8ea000, 28328 <unfinished ...> 32435 accept4(-1, 0xbfa8ccfc, [128], 0) = -1 EINVAL (Invalid argument) 32434 <... munmap resumed> ) = 0 32435 gettimeofday({1337707092, 785296}, NULL) = 0 32434 munmap(0xc70000, 81680 <unfinished ...> 32435 write(2, "[Tue May 22 19:18:12 2012] [erro"..., 100 <unfinished ...> 32434 <... munmap resumed> ) = 0 32435 <... write resumed> ) = 100
32437 open("/proc/sys/kernel/ngroups_max", O_RDONLY <unfinished ...> 32436 <... close resumed> ) = 0 32373 <... clone resumed> child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0xb78797b8) = 32439 32438 rt_sigaction(SIGUSR1, {0x5d45a0, [], SA_INTERRUPT}, <unfinished ...> 32437 <... open resumed> ) = 8 32436 munmap(0xb788d000, 4096 <unfinished ...> 32438 <... rt_sigaction resumed> {0x5d3d40, [HUP USR1], 0}, 8) = 0 32437 read(8, <unfinished ...> 32436 <... munmap resumed> ) = 0 32373 waitpid(-1, <unfinished ...> 32439 rt_sigaction(SIGHUP, {0x5d4f00, [], SA_INTERRUPT}, <unfinished ...> 32438 geteuid32( <unfinished ...> 32437 <... read resumed> "65536\n", 31) = 6 32436 setgroups32(1, [48] <unfinished ...> 32373 <... waitpid resumed> [{WIFEXITED(s) && WEXITSTATUS(s) == 1}], WNOHANG|WSTOPPED) = 32434 32439 <... rt_sigaction resumed> {0x5d3d40, [HUP USR1], 0}, 8) = 0 32438 <... geteuid32 resumed> ) = 0 32437 close(8 <unfinished ...> 32436 <... setgroups32 resumed> ) = 0
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
Murphy's Law is recursive. Washing your car to make it rain doesn't work.
On Tue, 2012-05-22 at 19:24 +0200, Luigi Rosa wrote:
$ strings /usr/lib/libapr-1.so.0|grep apr_socket_accept apr_socket_accept
Ok, the strings command gives the same result.
It should just for a pointer...
# Do not do a gracefull restart here! service httpd stop service hpptd
got a lot of Process 32565 attached
Should also strace all subprocesses off the parent
grep returns an empty set, there is no 'apr' in the entire httpd.txt Some (I suppose) relevant parts:
You want to look for errors also. You could post the entire thing to a pastebin that bit is basically the start up. If it is still doing it the error should be in there and may be socket related also as your was.
But be aware that some Parent Processes are not aware of there Subprocs and thus they can not be seen by the parent. So that entails the error will not be in the log!
Before you go any further:
httpd is running here lsof |grep httpd ... all process listed here
service httpd stop lsof|grep httpd ...should show no processes listed if you see any listed theres your problem maybe.
Your problem seems related to an older apache problem.
John
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Luigi Rosa said the following on 22/05/12 16:52:
Still getting the same error.
"solved" installing and configuring lighttpd which is now working like a charm
Still puzzled because the VM with this issue is a clone of a VM that I have at home which is working as expected. I cloned the same VM many times, but just that one has problem with apache.
Ciao, luigi
- -- / +--[Luigi Rosa]-- \
To see a need and wait to be asked, is to already refuse.
On Tue, 2012-05-22 at 19:02 +0200, Luigi Rosa wrote:
"solved" installing and configuring lighttpd which is now working like a charm
Still puzzled because the VM with this issue is a clone of a VM that I have at home which is working as expected. I cloned the same VM many times, but just that one has problem with apache.
Well now you have directions how fix it .... instead of installing an alternative so lets fix it:-)
On 05/22/2012 02:49 AM, Luigi Rosa wrote:
I have a VMware virtual machine with CentOS 6 32bit updated to the lates patches
Yesterday Apache started to give this error:
[Tue May 22 09:46:07 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:08 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket) [Tue May 22 09:46:09 2012] [error] (38)Function not implemented: apr_socket_accept: (client socket)
Any suggestion about this?
Did you update any packages ... we released a new nfs-utils package yesterday.