0 packet lost on my end. (But seems like my unifi prefix block get higher latency than yours)

3ms or 3000ms? Should be 3ms i think?
fdc routing not optimized for me,so i think it is normal.
Just now my area electricity maintainance so power down. Just power on my RPi 20min ago, now accessible already.

Oracle sg to unifi also not stable.
For azure it is because of the private peering between Microsoft and Unifi,so the connectivity is good.

You need to make sure your router support ipv6,and it might need to enable manually. (most tm provided router is supported)

Unifi to OVH SG start having latency spike at night...

Seems like today almost 98% network stability for daily hours.
Hopefully tonight can having better network than previous 5 months.

I think there was a short network outage at that time. From 5.12am to 5.20am.

no problem on my end-Johor. IPv6 Enabled.
So cloudflare and fast.com speedtest is using ipv6.
TM speedtest is IPv4.

Try reboot router to change a unifi ip prefix block?
And try speedtest.net to West Malaysia server(umobile KL and Digital Nasional Berhad) and East Malaysia server (Celcom Axiata Kuching and Celcom Axiata Kota Kinabalu) and Singapore server(prefer i3D.net) to see any abnormal to southeast asia server network.
If to east malaysia server normal but others not stable,then that might be some issue on TM or submarine cable between West and East malaysia.

Looks like is specific area issue,mine totally fine for any usage.

Started around 7pm

Information from Lowendtalk:
It's the AAE1 cable, multiple cuts and is likely to take a few days to fix.

no problem on my end,seems like occurs at specific region issue.

Some Edgecast CDN routing issue,some server is fast for malaysia.

ping to pbs.twimg.com

Pinging cs45.wac.edgecastcdn.net [117.18.237.70]
Ping statistics for 117.18.237.70:
Packets: Sent = 50, Received = 39, Lost = 11 (22% loss),
Approximate round trip times in milli-seconds:
Minimum = 57ms, Maximum = 61ms, Average = 59ms

Pinging cs672.wac.edgecastcdn.net [199.232.44.159]
Ping statistics for 199.232.44.159:
Packets: Sent = 28, Received = 28, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 7ms, Maximum = 23ms, Average = 8ms

You can try to add the [199.232.44.159] to hosts file to mitigate the issue.
199.232.44.159 pbs.twimg.com

Yes,just add the line and save.(might need to restart browser)

Today Unifi definitely got problem to Europe network.

Let see tonight got problem or not. (On my smokeping only show that yesterday night got some spikes, but so far is acceptable)

I think not only OVH is affected,seems like most connection to SG is having similar issue on my end.

I think the private peering is still existed, but the problem is occurs on TM Unifi end,from what i saw from my smokeping record- the problem just occurs immediately after 9pm.

Unifi network QoS or congestion at night causing the instability for most connection that going in and out of malaysia.
Most local usage is okay.

You can run a traceroute to the server that you are having issue and attach the result to complain to TM and MCMC.

My traceroute result to those ips range. I think the problem is on TM end,and can only be solve by TM...