Strange behavior with multiple virtual parameters - a bug?

Michael Ducharme mducharme at gmail.com
Fri Sep 15 20:24:43 EDT 2017


I have added three virtual parameters onto some devices, which show three
different IP addresses - the loopback address, a tunnel address, and the
dyndns address.

A few devices have blank loopback address and blank tunnel address (because
they have no loopback or tunnel) and they do have the dyndns address. These
devices are for some reason sometimes pulling the dyndns address into the
wrong field. Each time I refresh the device it disappears from one field
and reappears in another - moving back and forth between the "loopback
address" field, the "tunnel address" field, and the correct dyndns address
field.

I thought there was a bug in my code, but I have gone through everything
carefully and it looks like it is simply not possible for my code to have a
bug that would cause the dyndns address to get mixed up with the others the
way I have written it - the dyndns address is pulled by grabbing the serial
number from the device and using dns.lookup in an external module to get
the dyndns IP address.

The other two (the loopback and the tunnel address) are pulled directly
from the device via TR-069. Because the functions are so different, it is
impossible for the device to return the dyndns address to the virtual
parameter code for the loopback and tunnel address since it doesn't even
provide that address directly via the TR-069 stack.

Instead, it appears to be a GenieACS bug. It almost seems like GenieACS is
getting the return values mixed up. It doesn't happen with all devices
though - it only happens with these two that have only one out of three of
the parameters on the device, and that one is the dyndns. Devices that have
all three or two out of three (or none) display correctly.

Could the callback somehow be returning the result to the wrong item?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.genieacs.com/pipermail/users/attachments/20170915/348512d1/attachment.html>


More information about the Users mailing list