As with the Singaporean ones, packet loss is occurring at the *eqsy4.as57976.net points as well, and comes up reliably comes up within 10 mins, outside on raids and instances, even just standing around in Boralus.
So no, not an isolated issue with Starhub ISP in Singapore, but regionally where traffic goes through Singapore.
If you do WinMTRs of your own when a lag spike episode is in progress, you’ll see that all Malaysian traffic regardless of ISP (TM, Maxis, TIME) are routed to Australia via Singapore (sgw.equinix.com), and as soon as it hits Blizzard nodes at *.as57976.net, the data packets start dropping all along to the WoW server at 103.4.115.248.
It’s affecting us the same way as it’s affecting the Singaporeans.
And here’s another WinMTR from a lag spike that was in progress
As the routing is going through Singapore, the issue is likely going to be related to what’s going on in the thread over here. It is something we are looking into. In the meantime, using a VPN to alter the routing of the connection may be a potential workaround.
We made some modifications to the singapore routing recently that seems to have improved things. If you’re still having problems can you please follow up at this thread for me? I want to make sure all the reports are in one place. Thanks!