Thread Info
  • State Not Answered
  • Locked Locked
  • Replies 2 replies
  • Subscribers 27 subscribers
  • Views 4929 views
  • Users 0 members are here
Options
Suggested
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

WAN gateway goes down frequently. Always comes up when running diagnostics (ping, traceroute)

rgomez

Hello,

I have been experiencing a strange issue since July 8th: one of my 2 internet connections, a FO one, keeps going down frequently, but when I go into the admin page and try to run some diagnostics, say, ping 8.8.8.8 on that interface, then it comes up. If there are no diagnostics run then it usually comes up by itself at some point, some times minutes, sometimes hours. But as far as I have seen, if I go and do a ping on the diagnostics page it usually comes up right there.

I have already asked my ISP about this but, no surprises there, they say there is nothing wrong on their end. They asked me to run ping on some addresses when this happens, but as I said, doing that brings the connection back online. If I run the ping from outside the firewall, say, my computer, then it stays down. There is nothing needed to do on the OTN (I mean, no need to reboot or disconnect or something like that on that equipment), it doesn't display any error nor anything like that.

I have tried changing the monitoring parameters to include more addresses, doing TCP instead of PING, and so on, but nothing really helps. I have already lowered the MTU settings too but no real improvement either. As this is somewhat chaotic (I can't see any real pattern here) I can't evaluate if things are better or not. The only thing that I feel is that at night (when there is very low internet usage) it comes down more frequently than in the day. This would appear to be some kind of "disconnect because there is no usage", although there is *always* some usage.

The interface is setup as a DHCP gateway, the OTN is configured as a bridge by the ISP. The XG Firewall is running on a Hyper-V VM, the NICs are uniquely assigned to it (not shared with the OS nor anything like that).

I don't recall having anything special happen on my equipment or configuration on that day precisely, but before that day I didn't have any problem at all, except for when the connection really went down (the OTN saying there was LOS or something like that).

I have checked the dgd.log file and it displays that all the tests fail before declaring the gateway dead, but I don't really understand why is it that running the diagnostic ping makes it come alive every time.

Any ideas? Thanks in advance



This thread was automatically locked due to age.
Parents
  • 0

    Hi,

    are you running on Sophos hardware or your own?

    If it is your own check the power settings on NICs to make sure that sleep or low power is not enabled?

    Ian

  • 0 in reply to rfcat_vk

    Hello Ian,

    Thanks for the reply. The firewall is running on a Hyper-V VM. The NIC (is actually a 2-port NIC, one for this OTN and the other for a VSDL modem) does not have anything like that configured, and the problem only happens with this gateway, not the other one.

    I don't know if the OTN has maybe for some reason settings like that configured, there is nothing I can do to check or configure it in bridge mode, and reversing it to router has to be done by the ISP. I might try moving the port used (it has 4) but if the setting was indeed changed (by the ISP, I guess) then I would expect it to be changed in all the ports.

    Anyway, might worth a try.

    Today, for instance, it was working without problems since I posted the original message until about 1 hour ago. It came back when I did the first ping. In the results it shows that onle 3 of the 4 pings were responded, the first one being the missed one.

Reply
  • 0 in reply to rfcat_vk

    Hello Ian,

    Thanks for the reply. The firewall is running on a Hyper-V VM. The NIC (is actually a 2-port NIC, one for this OTN and the other for a VSDL modem) does not have anything like that configured, and the problem only happens with this gateway, not the other one.

    I don't know if the OTN has maybe for some reason settings like that configured, there is nothing I can do to check or configure it in bridge mode, and reversing it to router has to be done by the ISP. I might try moving the port used (it has 4) but if the setting was indeed changed (by the ISP, I guess) then I would expect it to be changed in all the ports.

    Anyway, might worth a try.

    Today, for instance, it was working without problems since I posted the original message until about 1 hour ago. It came back when I did the first ping. In the results it shows that onle 3 of the 4 pings were responded, the first one being the missed one.

Children
No Data