RoadRunnR Posted September 13, 2012 Report Share Posted September 13, 2012 Hi, It seems that runing snomONE on a system with both IPv4 and IPv6 address is broken. I'm no sure since when, I seem to remember that it worked in the first 4.5 version. With 1075 and 1090 the log file goes crazy with this: [0] 2012/09/13 11:00:49: TCP(IPv4):Could not bind socket to port 443 on IP 0.0.0.0 [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 443 for HTTP/HTTPS [0] 2012/09/13 11:00:49: (IPv6)Could not bind socket to port 443 on IP [::] [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 443 for HTTP/HTTPS [0] 2012/09/13 11:00:49: TCP(IPv4):Could not bind socket to port 389 on IP 0.0.0.0 [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 389 for HTTP/HTTPS [0] 2012/09/13 11:00:49: (IPv6)Could not bind socket to port 389 on IP [::] [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 389 for HTTP/HTTPS [0] 2012/09/13 11:00:59: (IPv6)Could not bind socket to port 80 on IP [::] It repeats over and over, connections to the webinterface are intermittend (since it shuts down port 80 and 443 and restarts them all the time). This also kills auto-provisioning. The only ports that seem to work reliably are 5060 and 5061 (SIP and RTP). Interface config: eth0 Link encap:Ethernet HWaddr 00:50:56:ae:17:48 inet addr:192.168.13.8 Bcast:192.168.13.255 Mask:255.255.255.0 inet6 addr: 2001:6f8:12d9:13::8/64 Scope:Global inet6 addr: 2001:6f8:12d9:13:250:56ff:feae:1748/64 Scope:Global inet6 addr: fe80::250:56ff:feae:1748/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:9843790 errors:0 dropped:0 overruns:0 frame:0 TX packets:8415647 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:2030992331 (2.0 GB) TX bytes:1692246693 (1.6 GB) lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:2995 errors:0 dropped:0 overruns:0 frame:0 TX packets:2995 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:369367 (369.3 KB) TX bytes:369367 (369.3 KB) System: # uname -a Linux vlx049-tpo 2.6.32-42-generic-pae #95-Ubuntu SMP Wed Jul 25 16:13:09 UTC 2012 i686 GNU/Linux Andreas Quote Link to comment Share on other sites More sharing options...
RoadRunnR Posted September 13, 2012 Author Report Share Posted September 13, 2012 Got it work after switching from the debian4 binary to the centos one. Seems the debian4 version was linked with an libc version that has problems with IPv4/v6 dual stack setups Quote Link to comment Share on other sites More sharing options...
pbx support Posted September 13, 2012 Report Share Posted September 13, 2012 Just FYI.. [0] 2012/09/13 11:00:49: TCP(IPv4):Could not bind socket to port 443 on IP 0.0.0.0 [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 443 for HTTP/HTTPS [0] 2012/09/13 11:00:49: (IPv6)Could not bind socket to port 443 on IP [::] [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 443 for HTTP/HTTPS [0] 2012/09/13 11:00:49: TCP(IPv4):Could not bind socket to port 389 on IP 0.0.0.0 [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 389 for HTTP/HTTPS [0] 2012/09/13 11:00:49: (IPv6)Could not bind socket to port 389 on IP [::] [0] 2012/09/13 11:00:49: FATAL: Could not open TCP port 389 for HTTP/HTTPS [0] 2012/09/13 11:00:59: (IPv6)Could not bind socket to port 80 on IP [::] This error will occur if the ports are being used by another service. That other service could previous instance of the snomONE PBX too. Under Linux systems, OS does not release the ports up to a minute after the process has exited. Quote Link to comment Share on other sites More sharing options...
Vodia PBX Posted September 13, 2012 Report Share Posted September 13, 2012 We dont link statically. There are some issues with binding to "all addresses", in order to make IPv6 migration easier some OS also bind to IPv6 addresses. We tell the OS no to do so, but maybe something really is different in the library that screws it up. Anyway, the /proc interface has a settings for that (/proc/sys/net/ipv6/bindv6only), you can try to put a "1" into that and then the PBX should always bind only to the IPv4 and Ipv6 interface seperately. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.