Apinger "detects" 100% packetloss every few days
-
Hi everyone,
I'm using 2.1-release on a PC Engines apu, behind a Fritz!Box (German Brand of DSL Router - pfS WAN with NAT = Exposed Host on Fritz!Box) which runs almost fine.
The only issue I have is apinger falsedetecting 100% paketloss every few days.
I've set the WAN gateway to ping Googles 8.8.8.8 which is also set as DNS, as it makes no sense to ping the gateway IP (which is the Fritz!Box and therefore would always answer as long as it is powered)
After I restart apinger everything runs smooth again.I'm not able to exchange the Fritz!Box for a simple modem as it also acts as the VoIP gateway for an ISDN System.
Is there a solution for that behavor?
Perhaps in 2.1.1?I'm a bit concerned to update, as I read there are some issues with it and the box is in production.
Any advice is highly apreciated.Thanks,
Harry -
Yes, apinger does report unusual %ages sometimes. I don't think that has been totally sorted in any version yet. But I do believe that 2.1.1 has loads of good fixes to all sorts of things, and no regressions that I am aware of.
Currently there is a big security issue about OpenSSL across the internet as a whole. pfSense 2.1.2-RELEASE is building right now. I would wait for 2.1.2-RELEASE to appear and then upgrade. (It should be in less than 24 hours, barring other stuff happening) -
Hi phil,
thanks for your infos.
It seems that the apinger problem is still present in the current release.
I've noticed, that it might have something to do with a (DSL/Cable) router beeing the WAN gateway, not a modem.
I have 3 pfS boxes and the one with the recurring packetloss issue is the only one connected behind a router.Could that really affect apinger?
Thanks again,
Harry -
I haven't worked out a good pattern for when this happens. I run most of my pfSense behind ADSL routers that have the real internet connection. So I have Alternate Monitor IP set on these. The numbers for %age packet loss and latency can go a bit off at times - I wish I knew what sequence of real packet loss and latency over time can cause the average calculations to go awry.
Now that pfSense-tools has been closed off, I can't see the details of the apinger source code mods for pfSense, so no chance to help find any problem. -
There is NOTHING the community could help with as long as the closed source apinger stuff is concerned. Go file a bug!
-
I have this happen as well on 2.1.5 but a reload of the GUI brings it back online….
Could it be widget related?
-
I have this happen as well on 2.1.5 but a reload of the GUI brings it back online….
Could it be widget related?
I have the same problem :-[
-
Same problem here, sometimes i get values over 100%. I have noticed that this used to happen on servers with WIMAX routers, this routers has a bigger trend to go down, i think that apinger goes crazy sometimes when a gateway goes down and up.
-
I only see it on WAN DHCP connections. Not on Static IP's at all.
It runs very smooth on that setup using 2.1.4.
My home setup has issues, but not the datacenters.
-
just adding my 2cents:
I have the inverse of Supermule:
- Home (dhcp) no issues at all
- Office (dhcp but static lease) apinger behaves quite odd.
Same isp/technology, so not sure it is to pin on the dhcp.
Both running 2.1.5.
-
I have turned on time sync between host and client on ESXi. I havent had the issue at all since last night
Oct 17 19:59:33 apinger: SIGHUP received, reloading configuration.
Oct 17 19:59:32 apinger: Starting Alarm Pinger, apinger(21363)
Oct 17 18:56:57 apinger: alarm canceled: WAN_DHCP(62.243.131.197) *** delay ***
Oct 17 18:56:49 apinger: ALARM: WAN_DHCP(62.243.131.197) *** delay *** -
My installation is completely static, currently on 2.1.5.
Interestingly, I've gone from having the problem every couple of days, to not having had the problem in 3 weeks. Probably because my connection has been very stable with little to no loss and no spikes of latency (which is what usually sets apinger off).
-
Goto
System -> Routing -> Gateways
choose Gateway and increase
Latency Threshold and
Package Loss Threshold
(maybe even "Down")
to e.g. 300-2000 ms and see if it helps….