Status
Not open for further replies.
It's looking better but I am still getting -some- weird packet loss and ping spikes.

Been having same issues since the weekend as well.

I just wish we can get to the bottom of this weird issue as online gaming has become a hit and miss lately. Sometimes ping is stable around 160-170 and then sometimes it jumps all over the place with packet drops every randomly throughout the session.

So many MTR's have been posted here now already that I feel we need to start getting some sort of answers. Is it overseas servers, is it Vuma, CISP...
 
Hi all, it looks like there was a brief issue with the LINX peering fabric last night. They did a firmware upgrade which affected all members on that switch. Please let me know if there are any issues still persisting.
 
Been having same issues since the weekend as well.

I just wish we can get to the bottom of this weird issue as online gaming has become a hit and miss lately. Sometimes ping is stable around 160-170 and then sometimes it jumps all over the place with packet drops every randomly throughout the session.

So many MTR's have been posted here now already that I feel we need to start getting some sort of answers. Is it overseas servers, is it Vuma, CISP...
What sort of issues have you been having since the weekend? Have you logged it?
 
What sort of issues have you been having since the weekend? Have you logged it?

Not logged a ticket, but surely with all the feedback from everybody here in the last couple of weeks with the same issues in gaming that must be enough to have a look at it?

I keep making screenshot records when it is bad so will post them tonight, but my problem is I always forget or when I get on at night, that particular night everything is fine. So ja, I just need to remember to post the screenshots.

That being said, something is not right, and there have been way too many people that have now come and complained about the same particular issues that are on the Vuma network.
 
Not logged a ticket, but surely with all the feedback from everybody here in the last couple of weeks with the same issues in gaming that must be enough to have a look at it?

I keep making screenshot records when it is bad so will post them tonight, but my problem is I always forget or when I get on at night, that particular night everything is fine. So ja, I just need to remember to post the screenshots.

That being said, something is not right, and there have been way too many people that have now come and complained about the same particular issues that are on the Vuma network.
Last night was a fault on the LINX peering fabric, as mentioned in my previous post. It hasn't happened before so unrelated. Please log it for me and attach as much info as you can.
 
so testing this a little further it seems to only effect 1080p downloads. Any other def comes down at max line speed. IDM allows one to select specific resolutions...

Scrap that... does it on 720p too...

Higher res full line speed to completion..
[MENTION=394133]PBCool[/MENTION] so same video, just different resolutions... Quite interesting...

[video=youtube_share;dhxpBzaHRQE]https://youtu.be/dhxpBzaHRQE[/video]

[MENTION=394133]PBCool[/MENTION] any explanation for this?

TIA
 
Was playing World of Warcraft (EU) last night, and my ping seemed fine. The issue was crazy lag spikes, and rubber banding. I haven't been playing long, so will have to see if it continues.
 
[MENTION=394133]PBCool[/MENTION] any explanation for this?

TIA

Kind of simple. You'll find that large resolution video's are probably saved to different servers. Also high resolution videos span more disk space and with that the possibility of more disks.

This all adds alot more that can slow it down too.

Things such as more people wanting to download higher resolution so the server is saturated or the load is higher on that server than it is for lower resolutions
 
Kind of simple. You'll find that large resolution video's are probably saved to different servers. Also high resolution videos span more disk space and with that the possibility of more disks.

This all adds alot more that can slow it down too.

Things such as more people wanting to download higher resolution so the server is saturated or the load is higher on that server than it is for lower resolutions

But the 4k video comes down at full line speed while the lower resolution videos (1080p and below) are consistently throttled to 4-5MB/s after an initial burst at full line speed.

I don't understand your server/disk theory either. Large resolution videos aren't the problem...
 
But the 4k video comes down at full line speed while the lower resolution videos (1080p and below) are consistently throttled to 4-5MB/s after an initial burst at full line speed.

I don't understand your server/disk theory either. Large resolution videos aren't the problem...

Not many people have the equipment to watch 4k video. So whats the point in downloading it at 4k?
You don't buy 17 inch rims for your 15 inch car tires xD

It's not all about size. It's about usage... If a server has say 100 mbps bandwidth as an example.

Server 1 is a high disk space server, so this is where the 4k videoes are stored. It has 5 people trying to download 4k video. effectively 100 / 5 = 20mbps theoretically for each person.

Server 2 is more mid range and stores 1080p and 720p. it has 20 people trying to download. effectively 100 / 20 = 5mbps

There is a whole lot more to do with load balancing and disk sharding and the likes, but thats a bit complicated to get into why it can speed up downloads and why it can slow them down.

Basically my point is, There are many many many reasons why the same video, at different resolutions (Effectively 2 completely different files according to whatever OS the servers are using) would download and be processed at different rates
 
Nah dude, I can replicate this behavior with monotonous regularity. The server load explanation doesn't cut it...

[video=youtube_share;y6IrHrOOLbg]https://youtu.be/y6IrHrOOLbg[/video]
 
Argh anyway, not going to pollute this thread with this any longer. It's not a deal breaker but certainly quite interesting to observe. Smells a lot like throttling and/or shaping! :p
 
[MENTION=394133]PBCool[/MENTION] any explanation for this?

TIA

I'm pretty sure I explained this already? We peer with Google locally and have a cluster on our network, why would we want to throttle this traffic?
 
I'm pretty sure I explained this already? We peer with Google locally and have a cluster on our network, why would we want to throttle this traffic?

yeah you explained that to me. Should have a general questions asked to answer things for people as this form is massive now
 
My Cacti server is telling me the same thing..
https://imgur.com/a/KJzwzNC
[MENTION=394133]PBCool[/MENTION], do you know why this might happen? I'm using the CISP DNS servers FWIW.

They load balance their content servers between JHB and Cape Town, they are apparently looking to change this in future but they arent sure when. It's an arbitrary way of doing it if you ask me, we also have to backhaul all of their traffic for them this way.
 
They load balance their content servers between JHB and Cape Town, they are apparently looking to change this in future but they arent sure when. It's an arbitrary way of doing it if you ask me, we also have to backhaul all of their traffic for them this way.

Is there a danger in pointing that hostname to the JHB IP on my router bypassing their load balancing?
 
Status
Not open for further replies.
Top
Sign up to the MyBroadband newsletter
X