On 01/15/2017 04:44 AM, Axel Beckert wrote:
Finally found a system with good old sysvinit and whose (sole) network
interface is named "enp0s31f6". So that's definitely not
systemd-specific.
I though don't have xymon-client installed on that machine. So if it
doesn't come out that the new net-tools output is the cause, I'll
probably have to set up a VM to test that.
Regards, Axel
Axel,
Thanks for the reply and for the information. I agree with you that it
is not systemd, but rather udev oriented. I was able to reconfigure the
wifi interface (orig: wlp4s0) to be wlan0 using an appropriate udev link
file last night.
Additionally, I downgraded the net-tools package to match the current
version for the jessie branch, and then let it run overnight.
In theory this should reduce the number of variables in play, however,
my results were surprising (nothing new generated) and that led to
further investigation.
What I'm seeing with new eyes is that while the msg.$HOST.txt file on
the client has the appropriate [ifstat] block in it with valid data, I'm
still not getting new RRD files, ifstat.*.rrd, generated on the xymon
server for display. The files are dated early this month.
This laptop that is having issues producing data is too new to run the
stock jessie kernel, so I used backports to insall the 4.6 kernel image,
but ultimately did a dist-upgrade to testing. This is where I found the
new net-tools display issue, but the interface naming was present at
install and kernel upgrade.
If you have some suggestions on how to force the generation of new RRD
files for the ifstat.*.rrd files, I'm happy to do that. I just haven't
figured out the xymon code enough to do that already.
david