Bug #4528
closed
no DynDNS RFC2136 Updates are done on some pfSense-installations
0%
Description
We have 4 installations of pfSense on KVM and configured a existing bind9 server with rfc2136 with your how-to (https://doc.pfsense.org/index.php/RFC2136_Dynamic_DNS).
On 2 installations everything works well - the WAN-Connection via PPPoE starts and the update is done.
On the other 2 installations no update is done - and on the bind9 Server with wireshark we don't even see a single packet coming.
We tried it with pfSense 2.15, 2.2 and 2.2.1, 2.1.5 updated to 2.2 and updated to 2.2.1. We tried to copy the VM from a working site and changed the settings. We tried also a 32bit 2.2 Version with fresh installation.
We tried also the TCP-switch, but nothing helped and we have either no or not actual log entries in System-General - e.g. we have this log:
Mar 9 10:40:39 GW syslogd: exiting on signal 15
Mar 9 10:40:39 GW check_reload_status: Syncing firewall
Mar 9 10:35:07 GW php-fpm33478: /services_rfc2136_edit.php: phpDynDNS: updating cache file /conf/dyndns_wan_rfc2136_'THE-HOSTNAME'_IP-OF-BIND9.cache: 188.110.221.241
Mar 9 10:35:07 GW check_reload_status: Syncing firewall
Mar 9 10:33:29 GW check_reload_status: Syncing firewall
Mar 9 10:32:58 GW php-fpm33478: /services_rfc2136_edit.php: phpDynDNS: updating cache file /conf/dyndns_wan_rfc2136_'THE-HOSTNAME'_IP-OF-BIND9.cache: 188.110.221.241
Mar 9 10:32:58 GW check_reload_status: Syncing firewall
Mar 9 10:30:10 GW php-fpm33478: /services_rfc2136_edit.php: phpDynDNS: updating cache file /conf/dyndns_wan_rfc2136_'THE-HOSTNAME'_IP-OF-BIND9.cache: 188.110.221.241
Mar 9 10:30:10 GW check_reload_status: Syncing firewall
Mar 9 10:27:51 GW check_reload_status: Syncing firewall
Mar 9 10:27:43 GW php-fpm26494: /services_rfc2136_edit.php: phpDynDNS: updating cache file /conf/dyndns_wan_rfc2136_'THE-HOSTNAME'_IP-OF-BIND9.cache: 188.110.221.241
Mar 9 10:27:43 GW check_reload_status: Syncing firewall
Mar 9 10:27:06 GW check_reload_status: Syncing firewall
Mar 9 10:26:08 GW php-fpm10648: /services_rfc2136_edit.php: phpDynDNS:updating cache file /conf/dyndns_wan_rfc2136_'THE-HOSTNAME'_IP-OF-BIND9.cache: 188.110.221.241
Mar 9 10:25:44 GW php-fpm242: /index.php: Successful login for user 'admin' from: 10.1.0.3
Mar 9 10:25:44 GW php-fpm242: /index.php: Successful login for user 'admin' from: 10.1.0.3
but the log was taken on 17. March - no single entry since a week, many reconnects and other ips since then and no update at all - at all means no update since installation.
We have logged on via ssh and send the nsupdate command per hand and it worked fine.
The .cache file is existing and contains the actual IP.
All machines have virtio network cards (Disable hardware checksum offload is set) and all other things are working well - many thanks for this great product.
What else could we try to provide you the right information to find (and eleminate) this bug....arrr..or our fault;-)
Updated by Chris Buechler over 10 years ago
- Status changed from New to Feedback
it definitely works in general. what's different about the ones that don't work vs. the ones that do?
Updated by Manuel Gräble over 10 years ago
The hardware of the proxmox VE servers is different, but all are on actual version of proxmox VE - 3 Installed with VE 3.3 and upgraded on non-subscription repository, 1 Installed long ago and upgraded with subscription repository - the one with the subscripted repository and on of the otheres is working.
Different Network Cards on the hosts:
Working Nodes:
1. r8169 for LAN and e1000e for WAN
2. e1000e for LAN and 8139too for WAN
not Working Nodes (have the same hardware):
1. r8169 for LAN and e1000e for WAN
2. e1000e for LAN and r8169 for WAN (We changed eth0 and eth1 for testing)
The not working Nodes only have a Intel J1900 Celeron Processor - the others a i7-920 and a Core 2 T6600 - but will the heardware matter for the pfSense as Virtual Machine? All pfSense Installations are working with virtio network cards - but we also tried an Installation with emulated Intel e1000 with same result.
The Internet Service Provider is also different - the working ones are at German Telekom and M-Net, the other ones at Arcor and 1&1, but the provider shouldn't matter, because the update by hand is working.
one working and the two not working installations have a Fritz!Box 3170 configured as DSL-Modem - the other working one has a standard Telekom DSL-Modem.
And, with copying a working VM to a not working node, changing the settings to fit the other network (Domain, IP, PPPoE, local DNS, ...) don't update also and no actual log entries are made.
By the way - on fresh installing 2.2 on x86 and x64 I wasn't able to change the keymap at the beginning of the setup, because I don't get entries for selection - same in Font and the third parameter (I don't remember at the moment - it's late today after 17h of work...)
Just bevor the Ticket, I upgraded one working and the two not updating Nodes via Webminterface to 2.2.1 - same result - one working, two not updating, no log entries - but on all VMs the IPv4-Address is green in the RFC2136 list.
We're working on this since two weeks, but I don't know what else can be done.
Is there a way to debug the update process directly or write a log or something, because the System logs in General don't get the expected entries.
Which process/script... is doing the update and how is it called?
Updated by Chris Buechler about 9 years ago
- Status changed from Feedback to Duplicate
- Affected Version deleted (
All)
duplicates #6357