MyRepublic fibre - Part 5

[Lehnsherr]

“Our technical support team is ready to assist you at +65 67171680, daily from 9 am to 2 am”

If support is very important. then just go with the one that gives u 24/7 support.

 

[Lehnsherr]

And again for about a min. Same wee hours of the night/morning and 1x per 24 hours. I've sent this on to MR.

TBF MR's customer support has been good. Don't have to wait long for call to be picked up. They are knowledgeable (relative to others), friendly and they do check back, It is a pity this small issue keeps popping up.

One min disconnect once per day is technically no biggie. But it does kill an online match whether gaming or watching live sports.

Feb 7 01:22:10 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 7 01:22:14 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 7 01:22:14 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 7 01:22:19 WAN Connection: WAN(0) link up.
Feb 7 01:22:19 rc_service: wanduck 2179:notify_rc restart_wan_if 0
Feb 7 01:22:28 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 7 01:22:31 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 7 01:22:31 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 7 01:22:33 acsd: eth4: COEX: downgraded chanspec 0x1908 to 0x100a: channel 3 used by exiting BSSs
Feb 7 01:22:34 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 7 01:22:37 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 7 01:22:37 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 7 01:22:40 WAN Connection: WAN(0) link up.
Feb 7 01:22:40 rc_service: wanduck 2179:notify_rc restart_wan_if 0
Feb 7 01:23:03 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 7 01:23:08 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 7 01:23:08 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 7 01:23:14 wlceventd: wlceventd_proc_event(645): eth4: Deauth_ind 10:CE:A9:FB:66:ED, status: 0, reason: Unspecified reason (1), rssi:0
Feb 7 01:23:14 wan: finish adding multi routes
Feb 7 01:23:14 ntp: start NTP update
Feb 7 01:23:14 miniupnpd[8883]: shutting down MiniUPnPd
Feb 7 01:23:14 miniupnpd: it is advised to use network interface name instead of 192.168.50.1/255.255.255.0
Feb 7 01:23:15 miniupnpd[7995]: HTTP listening on port 44524
Feb 7 01:23:15 rc_service: udhcpc_wan 7923:notify_rc stop_samba
Feb 7 01:23:15 kernel: Archer TCP Pure ACK Enabled
Feb 7 01:23:15 Samba Server: smb daemon is stopped
Feb 7 01:23:15 miniupnpd[7995]: Listening for NAT-PMP/PCP traffic on port 5351
Feb 7 01:23:16 rc_service: udhcpc_wan 7923:notify_rc start_samba
Feb 7 01:23:17 wan_up: Find a 2.5G port on [eth0]
Feb 7 01:23:18 ping_target_with_size: Ping test is complete.
Feb 7 01:23:20 ping_target_with_size: Successful to ping target(100.71.0.1) with size(79)
Feb 7 01:23:20 wan_up: Restart DDNS
Feb 7 01:23:20 dhcp client: bound 100.71.19.136/255.255.224.0 via 100.71.0.1 for 86400 seconds.
Feb 7 01:23:21 WAN Connection: WAN was restored.

I had some issues initially as well when went over to 3Gbps
At first I thought is the copper SFP+ overheating
so i change to just using the LAN port but still same
then i change cable and resolved the issue

After that i changed router, so i used back the copper SFP+ ----> no issues
so i got curious, change back to the original cable as well to use with the copper SFP+, no issues either

lol not sure why. like as if the connection need to "settle in"

 

[froztheart]

“Our technical support team is ready to assist you at +65 67171680, daily from 9 am to 2 am”

Ohh, my bad, the last time I recalled it was 24/7.
But MR pretty stable la most of the time for me, apart from those maintenance/extended downtime..

 

[junialum]

all I can say is that one of your Ethernet ports (the WAN port, by the looks of it) is going up and down somewhat frequently, and the corresponding WAN up/down events are likely only triggering when your router's connectivity checker (?) happens to be checking at the same time as your port being down

can't say for sure whether that's due to a hardware (ONT/router/cable) problem, or due to the ONT bringing the physical link up and down due to GPON reconfiguration, so perhaps you could also try isolating the problem to one of those things?

e.g. if the Ethernet link drops one or twice a minute in your logs, try plugging a computer in directly to the ONT and seeing if the same thing happens, and also router <> computer on eth0, etc

Thanks! Appreciate the suggestions.

WAN has been stable minus those 1 min disconnects happening around once a day.
- It is very noticeable when the 1 min disconnect happens - IOT devices go down, users devices disconnect, etc. Otherwise connection is rock solid. My home users are online most of the day and I will get complaints if there is any disconnect
- Anyways, have just tested (pinging google every 1/2 second) to confirm this.

MR called back (their support is really pretty good). On their end, their logs show connection is fine. They checked against my router logs and did not see disruptions on their end (bar the one time I power cycled the ONT). No reason to doubt them.

So that leaves ONT, Router. Just hard reset my 3x XT8s and see how things go. If issues persist, then I will test ONT and router WAN ports separately and directly. (Hook up one laptop each.)

Edit: ONT replacement resolved the issue.
https://forums.hardwarezone.com.sg/threads/myrepublic-fibre-part-5.4981088/post-155314698

 

[junialum]

I had some issues initially as well when went over to 3Gbps
At first I thought is the copper SFP+ overheating
so i change to just using the LAN port but still same
then i change cable and resolved the issue

After that i changed router, so i used back the copper SFP+ ----> no issues
so i got curious, change back to the original cable as well to use with the copper SFP+, no issues either

lol not sure why. like as if the connection need to "settle in"

Could be an issue of how well the ethernet cable was plugged in. I half suspect that is my issue. After my hard reset I had a lot of disconnects. (Which was very unlike my main issue - 1 disconnect per day.) Suspected it was because I had moved the router and the ONT around while doing the hard reset. Re-inserted the same (new; just changed 2 days back) cable and pushed both ends in hard - disconnects went away. Could be ONT and/or router WAN port(s), could be cable. Anyways, I used a high pressure air blower to clear both WAN ports and both ends of the cable of potential micro debris.

 

[dudumomo]

Anyone facing like very short momentary connection loss at late hours of the night / wee hours of the morning? I have experienced this on:

29 Jan 0029-0031hrs (2 mins)
29 Jan 2029-2030hrs (1 min)
29 Jan 2359-0000hrs (1 min)
30 Jan 2337-2338 (1 min)
31 Jan 0545-0546 (1 min)
31 Jan 2247-2248 (1 min)
1 Feb 2336-2337 (1 min)
2 Feb 2233-2234 (1 min)
4 Feb 0424-0425 (1 min)
5 Feb 0037-0038 (1 min)

Is this normal? I did not experience this with MR in my previous two homes. Only happened in my current home. (same router throughout 3x XT8) Full log: https://jumpshare.com/s/oG1phFRzG1S0mjY8TQdA

I've faced same issue this week, with some disconnection around 8pm or so. It happens probably 2-3 times this week.
I could not found the root cause if it was on my side or not.
I'm a pretty big user of bandwidth (server at home), so I thought it was some issue on my side. But not sure.

 

[junialum]

Happened again. Same thing 1 min disconnect. Gonna test the wan ports of both ONT and router.



Feb 8 03:36:56 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:01 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:01 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:37:07 WAN(0) Connection: WAN(0) link up.
Feb 8 03:37:07 rc_service: wanduck 1680:notify_rc restart_wan_if 0
Feb 8 03:37:16 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:18 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:18 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:37:22 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:24 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:24 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:37:27 WAN(0) Connection: WAN(0) link up.
Feb 8 03:37:27 rc_service: wanduck 1680:notify_rc restart_wan_if 0
Feb 8 03:37:53 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:57 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:57 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:38:01 wlceventd: wlceventd_proc_event(645): eth4: Deauth_ind 10:CE:A9:FB:66:ED, status: 0, reason: Unspecified reason (1), rssi:0
Feb 8 03:38:01 wan: finish adding multi routes
Feb 8 03:38:02 ntp: start NTP update
Feb 8 03:38:02 miniupnpd[10517]: shutting down MiniUPnPd
Feb 8 03:38:02 miniupnpd: it is advised to use network interface name instead of 192.168.50.1/255.255.255.0
Feb 8 03:38:02 miniupnpd[2692]: HTTP listening on port 44057
Feb 8 03:38:02 rc_service: udhcpc_wan 2616:notify_rc stop_samba
Feb 8 03:38:02 miniupnpd[2692]: Listening for NAT-PMP/PCP traffic on port 5351
Feb 8 03:38:02 kernel: Archer TCP Pure ACK Enabled
Feb 8 03:38:03 Samba Server: smb daemon is stopped
Feb 8 03:38:04 rc_service: udhcpc_wan 2616:notify_rc start_samba
Feb 8 03:38:05 wan_up: Find a 2.5G port on [eth0]
Feb 8 03:38:05 ping_target_with_size: Ping test is complete.
Feb 8 03:38:07 ping_target_with_size: Successful to ping target(100.71.0.1) with size(79)
Feb 8 03:38:07 wan_up: Restart DDNS
Feb 8 03:38:07 dhcp client: bound 100.71.28.54/255.255.224.0 via 100.71.0.1 for 86400 seconds.
Feb 8 03:38:08 WAN(0) Connection: WAN was restored.

Edit: ONT replacement resolved the issue.
https://forums.hardwarezone.com.sg/threads/myrepublic-fibre-part-5.4981088/post-155314698

 

[Singrish]

Ohh, my bad, the last time I recalled it was 24/7.
But MR pretty stable la most of the time for me, apart from those maintenance/extended downtime..

You just say only lai liao. 2 forumers posted about connection issues

 

[Lehnsherr]

Could be an issue of how well the ethernet cable was plugged in. I half suspect that is my issue. After my hard reset I had a lot of disconnects. (Which was very unlike my main issue - 1 disconnect per day.) Suspected it was because I had moved the router and the ONT around while doing the hard reset. Re-inserted the same (new; just changed 2 days back) cable and pushed both ends in hard - disconnects went away. Could be ONT and/or router WAN port(s), could be cable. Anyways, I used a high pressure air blower to clear both WAN ports and both ends of the cable of potential micro debris.

This could be one also. Cos the same thing we all have equipment wise is the ONT

 

[Lehnsherr]

You just say only lai liao. 2 forumers posted about connection issues

So what are you driving at? you are not the contributing to help or give some ideas to their problems.

What for you are criticizing a provider which you didn't sign up for?

 

[froztheart]

You just say only lai liao. 2 forumers posted about connection issues

Isoloted cases different from system wide downtime leh

 

[Singrish]

So what are you driving at? you are not the contributing to help or give some ideas to their problems.

What for you are criticizing a provider which you didn't sign up for?

Dont think too much. Not driving to Bt Batok or Punggol. Am still considering orange, green or purple & in the 3 threads in HWZ purple seems to have more ppl talking about connection issues so obviously im worried & asking if outages happen often.

 

[Lehnsherr]

Dont think too much. Not driving to Bt Batok or Punggol. Am still considering orange, green or purple & in the 3 threads in HWZ purple seems to have more ppl talking about connection issues so obviously im worried & asking if outages happen often.

When they started out (I am talking about 100Mbps days) - alot of issues
Then it died down gradually, with people complaining still but it is mostly isolated cases
big ones - like whole area everybody kanna - at most once a year on average I feel

Right now its under SH, It is also said that after SH took over, users start to have more issues.

These are what I have observed over the years.

Bottomline is If you fear outage cos you depend on it for work or studies, you should have another connection (mobile so much data now or sign another fiber line). Cos sweeping outage they cant solve immediately, sticky problems will drag u for days or weeks. either way having another connection is the way out.

 

[cyberet]

Could be an issue of how well the ethernet cable was plugged in. I half suspect that is my issue. After my hard reset I had a lot of disconnects. (Which was very unlike my main issue - 1 disconnect per day.) Suspected it was because I had moved the router and the ONT around while doing the hard reset. Re-inserted the same (new; just changed 2 days back) cable and pushed both ends in hard - disconnects went away. Could be ONT and/or router WAN port(s), could be cable. Anyways, I used a high pressure air blower to clear both WAN ports and both ends of the cable of potential micro debris.

whats the cpu/2.4ghz/5ghz temperture? are you using merlin's firmware?

 

[junialum]

whats the cpu/2.4ghz/5ghz temperture? are you using merlin's firmware?

Not using Merlin. How do I get the temperature values?

 

[junialum]

Happened again. Same thing 1 min disconnect. Gonna test the wan ports of both ONT and router.



Feb 8 03:36:56 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:01 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:01 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:37:07 WAN(0) Connection: WAN(0) link up.
Feb 8 03:37:07 rc_service: wanduck 1680:notify_rc restart_wan_if 0
Feb 8 03:37:16 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:18 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:18 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:37:22 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:24 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:24 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:37:27 WAN(0) Connection: WAN(0) link up.
Feb 8 03:37:27 rc_service: wanduck 1680:notify_rc restart_wan_if 0
Feb 8 03:37:53 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link DOWN.
Feb 8 03:37:57 kernel: eth0: sysport_tm port shaper set to 999000 kbps, 35000 mbs (phy speed 1000000 kbps)
Feb 8 03:37:57 kernel: eth0 (Int switch port: 1) (Logical Port: 1) (phyId: 7) Link Up at 1000 mbps full duplex
Feb 8 03:38:01 wlceventd: wlceventd_proc_event(645): eth4: Deauth_ind 10:CE:A9:FB:66:ED, status: 0, reason: Unspecified reason (1), rssi:0
Feb 8 03:38:01 wan: finish adding multi routes
Feb 8 03:38:02 ntp: start NTP update
Feb 8 03:38:02 miniupnpd[10517]: shutting down MiniUPnPd
Feb 8 03:38:02 miniupnpd: it is advised to use network interface name instead of 192.168.50.1/255.255.255.0
Feb 8 03:38:02 miniupnpd[2692]: HTTP listening on port 44057
Feb 8 03:38:02 rc_service: udhcpc_wan 2616:notify_rc stop_samba
Feb 8 03:38:02 miniupnpd[2692]: Listening for NAT-PMP/PCP traffic on port 5351
Feb 8 03:38:02 kernel: Archer TCP Pure ACK Enabled
Feb 8 03:38:03 Samba Server: smb daemon is stopped
Feb 8 03:38:04 rc_service: udhcpc_wan 2616:notify_rc start_samba
Feb 8 03:38:05 wan_up: Find a 2.5G port on [eth0]
Feb 8 03:38:05 ping_target_with_size: Ping test is complete.
Feb 8 03:38:07 ping_target_with_size: Successful to ping target(100.71.0.1) with size(79)
Feb 8 03:38:07 wan_up: Restart DDNS
Feb 8 03:38:07 dhcp client: bound 100.71.28.54/255.255.224.0 via 100.71.0.1 for 86400 seconds.
Feb 8 03:38:08 WAN(0) Connection: WAN was restored.

36 hours of testing using PingPlotter. Laptop connected directly to ONT; second laptop testing WAN port of router. Results -
ONT: Momentary ping drop for several websites at the same time. 4 drops within a couple of seconds. Otherwise ok.
Router: Zero ping drops.

Could it be that the WAN port of the ONT has issues? This could be why MR backend does not see errors on my connection (optical signal is ok) but I am facing issues even though my router is fine. (I am not sure how ONTs work so not sure if this logic is sound.)

Will report this back to MR.

Edit: ONT replacement resolved the issue.
https://forums.hardwarezone.com.sg/threads/myrepublic-fibre-part-5.4981088/post-155314698

 

[cyberet]

Not using Merlin. How do I get the temperature values?

at the tools section. not sure if the stock firmware has this.

 

[junialum]

at the tools section. not sure if the stock firmware has this.

Ah ic. Stock does not have this.

 

[xiaofan]

whats the cpu/2.4ghz/5ghz temperture? are you using merlin's firmware?

Not using Merlin. How do I get the temperature values?

Ah ic. Stock does not have this.

You can use SSH to log in to the router and then issue the following commands to get the temperature. This will work for Asus routers using Broadcom chipset. For CPU temperature, it should work for Asus routers using non-Broadcom chipset as well.

https://forums.hardwarezone.com.sg/threads/asus-zenwifi-bt10-discussion.7067525/page-9

For more precise reading, you can SSH into the router.
https://www.snbforums.com/threads/how-to-read-the-temperature-without-merlin.80167/

CPU temp:

cat /sys/class/thermal/thermal_zone0/temp | awk '{print $1 / 1000}'

WiFi temp: for Broadcom chipset

for radioIF in $(nvram get wl_ifnames)
do
   wl -i $radioIF status | grep Chanspec | cut -d ' ' -f 2-5
   wl -i $radioIF phy_tempsense | awk '{print $1/2 + 20}'
   echo
done

Example run log for my Asus RT-AX86U. The CPU Core temp is high. WiFi PHY is not that hot.

xiaofan@RT-AX86U-2D40:/tmp/home/root# cat /sys/class/thermal/thermal_zone0/temp | awk '{print $1 / 1000}'
82.685

xiaofan@RT-AX86U-2D40:/tmp/home/root# for radioIF in $(nvram get wl_ifnames)
> do
>    wl -i $radioIF status | grep Chanspec | cut -d ' ' -f 2-5
>    wl -i $radioIF phy_tempsense | awk '{print $1/2 + 20}'
>    echo
> done
2.4GHz channel 9 20MHz
52

5GHz channel 155 80MHz
53

For TUF-BE6500, wl command does not exit as it uses Qualcomm IPQ5322 CPU. I can only get CPU core temperature.

xiaofan@TUF_6500-5020:/tmp/home/root# cat /sys/class/thermal/thermal_zone0/temp | awk '{print $1 / 1000}'
71.3

 

[junialum]

36 hours of testing using PingPlotter. Laptop connected directly to ONT; second laptop testing WAN port of router. Results -
ONT: Momentary ping drop for several websites at the same time. 4 drops within a couple of seconds. Otherwise ok.
Router: Zero ping drops.

Could it be that the WAN port of the ONT has issues? This could be why MR backend does not see errors on my connection (optical signal is ok) but I am facing issues even though my router is fine. (I am not sure how ONTs work so not sure if this logic is sound.)

Will report this back to MR.

MR will be changing the ONT. Anyways thanks everyone for helping with ideas, etc.

Edit: Wanted to add that info in this forum has been incredibly useful. Not just this thread but also other threads on choosing ISP, network equipment, etc. I am eventually going to move to 10gbps (currently on 1gbps) so getting up to speed on all components required (incl things like 10gbe NICs running hot) is helpful. Thanks again to the many contributors for keeping this community effort going.

Edit2: ONT replacement resolved the issue.
https://forums.hardwarezone.com.sg/threads/myrepublic-fibre-part-5.4981088/post-155314698