This is a weird disconnection problem that started on March 25. Can anyone offer advice trying to diagnose it?
I've had a home hub 5 operating perfectly for 13 months but ever since March 25 it's been disconnecting twice every night, an hour or two both before and after midnight. The commonest times are around 10pm and 1am. It never disconnects earlier in the day.
The orange 'b' icon appears and disappears periodically, but never goes blue by itself, even after I reset the hub, either using the reset switch or via the gui. However, either power cycling the hub or temporarily disconnecting the dsl cable from the wall enables an automatic reconnect with no other assistance.
I put my previous home hub 4 plus white modem box back for a couple of days and it did exactly the same thing. Off around 10pm ish and again at 1am ish. With this combination I had to power cycle the white box or temporarily disconnect the dsl cable to get a reconnect. I put the HH5 back and it disconnects as before.
Examples from the event log just before each disconnection are:
22:10:23, 25 Mar.
(2687434.990000) PPP LCP Send Termination Request [Peer not responding]
22:06:36, 26 Mar.
(85964.850000) PPP LCP Send Termination Request [Peer not responding]
02:38:18, 27 Mar.
(12123.810000) PPP LCP Send Termination Request [Peer not responding]
21:47:17, 27 Mar.
(37815.600000) PPP LCP Send Termination Request [Peer not responding]
23:33:34, 03 Apr.
(82637.290000) PPPoE is down after 1369 minutes uptime [Disconnected]
23:33:32, 03 Apr.
(82635.230000) PPP LCP Send Termination Request [Peer not responding]
00:59:50, 04 Apr.
(87812.860000) PPPoE is down after 84 minutes uptime [Disconnected]
00:59:48, 04 Apr.
(87810.770000) PPP LCP Send Termination Request [Peer not responding]
00:37:00, 04 Apr.
(86442.690000) CWMP: session completed successfully
00:37:00, 04 Apr.
(86442.490000) CWMP: HTTP authentication success from https://pbthdm.bt.mo
00:36:59, 04 Apr.
(86441.530000) CWMP: Server URL: https://pbthdm.bt.mo; Connecting as user: ACS username
I also see suspicous events like these sporadically, at various times of the day:
IN: BLOCK  Remote administration...
OUT: BLOCK  ICMP replay (ICMP type 3 code 1...
OUT: BLOCK  First packet is Invalid (Packet not in tcp window: TCP...
OUT: BLOCK  First packet is Invalid (Invalid tcp flags for current tcp state: TCP
and lots of wifi disconnected events at various times of the day like these:
(23484.540000) Device disconnected: Hostname: BT-DualBand-WiFi-Extender...
(23484.550000) Device disconnected: Hostname: braviaf2012...
(24304.560000) Device disconnected: Hostname: EPSON37ABD7
They don't seem to be related to the late night broadband disconnections.
Thanks for any advice.
Solved! Go to Solution.
did you also do a factory reset using the recessed button on back and hold till light flash?
did you check for any line noise4? dial 17070 option 2 should be quiet and best with corded phone
Is this another HHBa rebooting issue like the 1 I had?
I rarely see posts with HHB5b problems
I did check for a quiet line already, it's silent.
I'll do a factory reset now and report back tomorrow, thanks.
At 9:15pm I backed up the HH5, did a factory reset and restored the config, which forced a reboot. Good news, no disconnection at the usual time 10pm - 2am after that, but since it rebooted I think another night is needed to conirm if it has helped. I'll report back tomorrow...
Unfortunately the hub disconnected again tonight at 9:17pm so the factory reset didn't help.
It's clearly not the HH5 as I replaced it with my previous HH4 and that disconnects late at night in exactly the same curious way. What's the best way to report this as a fault to BT?
I do see pages and page of events throught the day for wired and wireless devices on the network that are currently switched off:
(83178.850000) Device disconnected: Hostname:<some device name>
Every five minutes or so, I see a list of events like this for every wired and wifi device that's switched off. What could cause this?
For anyone following this...
After 16 days of two guaranteed disconnections every night, an OpenReach engineer was seen working in our cabinet on Sat Apr 9. We've had no disconnections since.