Hi All,
This has been occurring for at least a year but could be localised to the chatswood CMTS. Would be interested to know if others experience similar issues with latency dependent services
During peak times of the day (7pm to midnight) we get drastic latency spikes on the cable connection. Jumps from the standard 8 - 14 ms first hop to above 500ms, sometimes multiple seconds. While this is annoying it only has a minor impact to general browsing. My housemate is an avid League of Legends player along with me occasionally playing Battlefield. These issues make it extremely frustrating to play or use any latency dependent applocations during these hours (It's probably made worse by these apps using UDP, Google services using quik are adversely effected compared to the rest of the net).
I have numerous MTR results (to be taken with grain of salt) and TCPDUMPS proving these issues.
I've ruled out any issues with our equipment as we have a Telstra CG3400D bridged with a Juniper SRX in conjunction with a backup internode business connection (From Fritzbox ) which has never had these problems all being routed through the SRX.
I typically just use the internode connection but Telstra has a 15ms faster round trip and prefer it for the obvious speed advantage.
Another possible issue impacting League of Legends / Riot services is Telstra's use of a tunnel or route optimisation of some kind, this seems to occasionally impact us but is unrelated to the localised issues impacting all connections.
rio1370642.lnk.telstra.net - Could any tech from telstra comment on what this is, I've confirmed its problematic from work Enterprise Telstra links that terminate in EQX (primary) and Globalswitch (secondary) (both 10gbit with sub 3ms latency)
Route is as follows, IP block is 60.225.18.128/18
1. 10.0.0.1
2. 10.204.192.1
3. 58.160.11.2
4. 58.160.15.226
5. bundle-ether4.ken-edge902.sydney.telstra.net
Please advise, I've been avoiding ringing support as getting this diagnosed through typical means will be frustrating.