Cool Ideas Fibre ISP – Feedback Thread 5

[CoolEscalator]

2 Pages back you can see their IPERF results from what i understand there is packet loss

Thanks, yeah there is heavy loss on those upload Iperfs, would recommend they share those results with Support as well.

 

[PBCool]

Can't speak for other things, but YT is dropping out every few minutes and then returning. Now my connection just completely dropped and came back multiple times. Additionally upload performance is still completely in the gutter. You have a network issue that began late last week, please stop pussyfooting around on it, it's starting to get exceptionally frustrating.

Hi, as mentioned, have you tried to bypass your router as yet? We have no current network incidents. I did mention Google are performing random maintenance at the moment.

 

[PBCool]

Oh wow that is not good

Change has been confirmed with Openserve now, next Tuesday apparently.

 

[Fulcrum29]

Hi, we have received 2 more reports since your message, Including Gordons bay and Strand.
I reached out to FF, and they have confirmed there are nodes down, and they will be logging this

EDIT:

Outage Reference	FRG4166415
Outage Title	Outage | CPT | Somerset Ring | Down
Region	Cape Town
Area	Blue Downs, Somerset West
Precincts Affected
Incident Start Date	2023-11-20 13:32:00.0
Outage Status	Resolution In-Progress

I have a idea that Frogfoot needs to upgrade their infrastructure for this area. There are some telltale signs of congestion.

 

[CoolEscalator]

I have a idea that Frogfoot needs to upgrade their infrastructure for this area. There are some telltale signs of congestion.

Would you be able to monitor and report this to our support, should any issues arise after the outage resolution?

 

[Tharaxis]

Speedtest with direct connection, bypassing the Ubiquiti network produces same results:



iPerf tests show 30-32% packet loss

It's at the worst levels it's been.

 

[PBCool]

Speedtest with direct connection, bypassing the Ubiquiti network produces same results:

View attachment 1620999

iPerf tests show 30-32% packet loss

It's at the worst levels it's been.

Cool please log all of that for me and the team will take it up with the fibre provider.

 

[Fulcrum29]

Would you be able to monitor and report this to our support, should any issues arise after the outage resolution?

Yes, I will check after my services are restored and if my issues persist. There is still no ETA for the current outage, hearing this from some very frustrated people.

 

[CoolEscalator]

Yes, I will check after my services are restored and if my issues persist. There is still no ETA for the current outage, hearing this from some very frustrated people.

Much appreciated.

Feedback below:

EDIT at 17:44pm @Fulcrum29
Outage Resolved

Outage Title: Outage | CPT | Somerset Ring | Down
Outage Ref(solid): FRG4166415
Impact Start Time: 2023-11-20 13:32:00
Impact Stop Time: 2023-11-20 16:57:38
Impact: Hard Down

Outage Update:
2023-11-20
-- Logged to FF Maintenance and Layer 1 vendor
-- Services affected by dual backhaul failure in the region
-- Techs have been dispatched

Link 1: Gbaynorth <> Goldenacre
-- Disconnection tested, layer 1 vendor are on site busy localizing
-- Vendor techs are busy prepping cable, possible ETTR is 16:35
-- Vendor has completed splicing however links remain down, FF maintenance team is on route golden acre to retest the link

Link 2: Bluedowns <> Acorncreek
-- Techs found high levels in the node and are busy investigating
-- Vendor techs are still busy troubleshooting, FF Maintenance team are en route to assist with investigation
-- Link has restored after swapping out the SFP, monitoring for stability
-- Services remain stable, outage resolved

 

[giggity]

Yeah looks like WACs express came back around the 25th of October.

I've noticed the past few evenings that latency to League of Legends EUW has been fluctuating between 173ms and 179ms again. Did WACs express path go offline again?

 

[CrypticZA]

I've noticed the past few evenings that latency to League of Legends EUW has been fluctuating between 173ms and 179ms again. Did WACs express path go offline again?

That would be load balancing between the two cables one from CT is 140ms and the other 146ms so there is the 6ms difference

 

[giggity]

That would be load balancing between the two cables one from CT is 140ms and the other 146ms so there is the 6ms difference

Yeah that explains the behaviour pretty perfectly. I get load balancing for high traffic like streaming, but it isn't great for gaming.

Any way Cool Ideas can pin connections? The game uses UDP, which I'm guessing would complicate it. Or is there any workaround like using a specific VPN?

Whenever it swaps over, there is some packet loss. I'd rather have constant 180ms than have it swap every minute or two to 173ms and have the game freak out for a split second each time.

 

[PBCool]

That would be load balancing between the two cables one from CT is 140ms and the other 146ms so there is the 6ms difference

This was resolved yesterday btw

 

[CrypticZA]

This was resolved yesterday btw

Ah so are both cables at 140 now?

EDIT* saw response below

 

[PBCool]

Yeah that explains the behaviour pretty perfectly. I get load balancing for high traffic like streaming, but it isn't great for gaming.

Any way Cool Ideas can pin connections? The game uses UDP, which I'm guessing would complicate it. Or is there any workaround like using a specific VPN?

Whenever it swaps over, there is some packet loss. I'd rather have constant 180ms than have it swap every minute or two to 173ms and have the game freak out for a split second each time.

It was resolved yesterday, was due to a break on Equiano path between Lisbon and London. Variance is now ~2ms which is back to normal.

 

[dillydog]

Hello sorry for the late message. Sure, Web Africa, 50mb = R649, 100mb = R789, Afrihost 50mb= R697, 100mb = R897, Clear Access ISP, 50mb = R699, 100mb = R899 ,Cool Ideas, 50mb =R809, 100mb = R999

Any feedback on this?

 

[leppie]

Youtube has been giving bad performance for a while now for everyone just look at the Afrihost thread Google even did maintenance on their cache's which seemed to make things worse

Yeah, getting worse and worse. Basically impossible to play high res video's anymore.

Today it is struggling even with 720p

 

[PBCool]

Yeah, getting worse and worse. Basically impossible to play high res video's anymore.

Today it is struggling even with 720p

Any examples? Are you using our DNS servers?

 

[leppie]

Any examples? Are you using our DNS servers?

Grab any 6 year old video from here https://www.youtube.com/@BILSTEINde/videos

Drops to 720p (or lower) instantly and 1080p is basically takes minutes to buffer.



Yes, using what you supply the router.

Edit:

Pulling from https://rr2---sn-jpn215-woce.googlevideo.com/videoplayback
Which resolves to 154.0.9.173

Tracing route to 154.0.9.173 over a maximum of 30 hops

  1    <1 ms    <1 ms    <1 ms  RT-AX55-DEA8 [192.168.0.1]
  2     6 ms     2 ms     2 ms  154.0.0.246
  3     2 ms     2 ms     2 ms  100.98.0.3
  4    19 ms    20 ms    19 ms  100.99.2.225
  5    19 ms    22 ms    18 ms  100.99.1.250
  6    19 ms    19 ms    19 ms  100.99.0.55
  7    19 ms    19 ms    19 ms  154.0.9.173

Trace complete.

 

[prodigyX]

Trying to update Diablo 4 through Blizzard's launcher, it spikes up to full speed for one second and then drops super low, stays there, then times out. Whats going on?

500/250 on openserv