Intermittent LAN drops...

faxfan · 19 Mar 2017 at 23:18

Guys,

Getting a problem on my network between a Draytek 2860 and a Cisco 3560CX-12PC, the logs I can get is from the switch when the connection drops for a about 3 seconds -

Mar 19 22:34:45.087: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet0/1, changed state to down
Mar 19 22:34:46.091: %LINK-3-UPDOWN: Interface GigabitEthernet0/1, changed state to down
Mar 19 22:34:48.726: %LINK-3-UPDOWN: Interface GigabitEthernet0/1, changed state to up
Mar 19 22:34:49.726: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet0/1, changed state to up

Every time is it for 2-3 seconds and it comes backup. Every couple of hours - though sometimes up to a 4 hours gap.

I'm assuming it's the not the cable - it is a 100ft cable run so not easy try another and it passes the cable test - and I'd see a more consistent problem.

I beginning to suspect the Draytek - seeing very high CPU peaks i.e. 3% jumping to 70%-80%, but I can't seem to easily log these.

Any ideas where to start on this?

bremen1874 · 20 Mar 2017 at 02:50

I'd connect another switch that can be monitored between the DrayTek and the Cisco at the DrayTek end of the 100ft cable. See if it still drops and which link it is.

Then move the switch to the other end of the 100ft cable and repeat.

faxfan · 20 Mar 2017 at 09:53

I know which link it is, just can't figure out why. But I think your right just need to sort a switch that can be monitored.

bledd · 20 Mar 2017 at 12:42

-Make a temporary cable and run it along the floor, see if that drops.
-Replace the wall module on end of the cable and re-punch the patch panel end.

Latest firmware for Draytek and Cisco?

APM · 20 Mar 2017 at 18:11

Have you tried a re-boot of all the hardware?

faxfan · 21 Mar 2017 at 12:22

Well that was an interesting one.

I connected another managed switch between the Cisco and Draytek. The connection remained stable but I still lost connections in / out of the LAN, using ECMO Ping Monitor (free ping monitor for 5 hosts) it showed the whilst DSL remained up I couldn't ping the nearest hop.

Called up ISP (ZEN) who could see PPP drops and suggested connecting to the master socket / new filter and everything is stable, though I suspect they may have done tweaking as well but I'll start reversing the changes.

So even though I was seeing a dropped connection on a LAN switch the actual problem was PPP drops to the ISP, obviously.