High ping to ACT WAN gateway

  • Thread starter Thread starter Smallship
  • Start date Start date
  • Replies Replies 20
  • Views Views 5,922
Messages
158
Location
Tamil Nadu
ISP
Airtel Black | 200 Mbps
I have taken a connection in Madurai, and they have assigned me a Chennai server as the WAN gateway. Usually the ping is 14-15ms to Chennai, but sometimes it goes to like 27-30 and stays like that for a day or two. Tried contacting ACT but they tell me to get a static IP. I tried explaining them that it's their gateway server(next hop after router local gateway) but they don't seem to understand. How should I proceed?
 
Can anyone help?
 
Not really, the increased latency to the gateway(second hop) directly adds up to all the servers. Sometimes it's 14ms sometimes it's 30. The difference adds up to all servers I connect to.
I'm asking them to fix the pings to the WAN gateway itself. The second hop. That's Problem X.
I used GoogleDNS to compare ping times to a Chennai server with the other ISP.
 
Okay. What are you trying to do that +30ms is bothering you? "Fixing" this would mean somewhere between huge amount of investment into their infrastructure to still significant amount of investment depending upon the actual problem. I don't think anyone is going to do that for one customer. Maybe you should take their offer on static IP and see if it works out.
 


The difference matters very much in competitive FPS. +5ms on average is definitely tolerable, but +20ms or above isn't. Plus the issue isn't just exclusive to me. I've checked with three other people around me who use ACT and they get similar ping times, which indicates a problem on their infrastructure that they need to fix. I don't require a static IP atm, but I also don't see how would it make a big difference in ping times. As a customer I'd expect lower pings from day one from how they advertise about their speeds and low latency. I should not need to pay more for basic elements like pings and speed, even if static IP could fix it for that matter.
 
Last edited:
Right. I don't see how their advertisement is wrong, 20ms to Google is extremely low. Still I don't think according to basic laws of Physics and Economics this something that they can solve for you tomorrow, so you should look into the other ISP that works for you or move somewhere where 2ms to Google is a thing.
 
I think you got me wrong. I put out the Google pings to compare with my backup connection to show the current difference, since I cannot ping my internal ACT gateway from another connection(my Gateway and Google IPs are in the same city, so the pings are the exact same).

The lowest possible I get is 8ms to Google on the other connection. While the lowest I get on ACT is 18ms and it even reaches 30ms(because the pings to gateway fluctuate to a higher number and stays like that for a day or two).

I don't expect the lowest pings like 8ms but at least it should be reasonable and stable. I wouldn't mind a 15ms ping to the gateway and Google. 30ms to WAN gateway is too high and it occurs randomly and takes many hours to get back to normal.

And yeah, I do completely understand the chances are extremely low that they fix it immediately, but still I'm trying to bring it to their notice.
 
Last edited:
Fluctuation in ping could very well be because of something on your end rather than ISPs. I'd say it's probably because of something on your end. Start by looking at thermal throttling.
 
I switch off the router at night and in some mornings I could get any ping to gateway/speedtest nearest server. 25-30 or 19ms. Jitter is <2ms most of the time. Restarting/switching off router for some time don't work either. It spikes to 30ms randomly, goes back to 18ms randomly, like after 3-4 hours, it doesn't matter if my equipment is on or off, for the ping to go up or down. It's unpredictable.

I think I have checked everything in my end. My router's pings, optical module stats, temperature and signal levels and ran ping tests in different devices, both ethernet and WiFi. I don't think anything is wrong in my end.
 
Last edited:
It was 21ms an hour back. It has come down to 16ms and I'd say this is fine. But it would go back to like 25ms or above randomly again. I'll post again when it does

This is the lowest I ever got, I guess. I would not complain even if it's 16-18ms.
 
The customer care explains that they use different paths and they have no control over it.
Kind of obvious they use the cheapest route. Having no data center in Madurai(connects to Chennai directly, which gives 30ms or more for a distance of 500km, just to connect to their servers.) the pings to any server is just terrible. Looking to switch to Xstream once it's available here. The experience with ACT is just disgusting. This has happened more than three times since I last posted. Takes days to go back to 20ms

Source
 
This is an issue I doubt any ISP would or be willing to solve.
There must be some disparities along the route to gateway (faulty cable or maybe they are just using copper), since you are getting the speeds paid for I highly doubt there's a weak edge packet forwarder on their end.
Does not seem to be something static IP would solve.
A lot of ISP use QoS and keep ICMP quite low on their priority list, which can also explain they delay. ICMP is not really a good way to detect RTT.
If it's really bugging I'd suggest you are better off trying a different ISP.
 

Top