Web Squad ISP

[Tokolosk]

Update on Googlegate: After hours of back and forth with multiple teams at Google, they still haven't figured it out. We found a pretty simple solution though - Google has 2 x 100G peering points at NAP Africa JHB, and simply turning down our sessions with one of these resolved any and all Google related issues. Things are looking much better since yesterday, so happy YouTubing and Google Playing!

Fantastic! Thanks!

 

[DeatheCore]

Update on Googlegate: After hours of back and forth with multiple teams at Google, they still haven't figured it out. We found a pretty simple solution though - Google has 2 x 100G peering points at NAP Africa JHB, and simply turning down our sessions with one of these resolved any and all Google related issues. Things are looking much better since yesterday, so happy YouTubing and Google Playing!

Just caught up on this thread. Funny thing is I noticed some strange buffering issues with YouTube a few days ago on my CISP connection in CPT too! Figured it was on Google's end so just ignored it. Issue seems to have gone away!

 

[MDE]

https://twitter.com/i/web/status/1225387510748061697

 

[Chiller89]

Hey @websquadza is there something weird going on internationally?
Keep getting random no route to host responses running a ping to s3 (trying to work on infrastructure hosted in EU)

64 bytes from 52.218.20.116: icmp_seq=425 ttl=34 time=201.349 ms
ping: sendto: No route to host
ping: sendto: No route to host
ping: sendto: No route to host
Request timeout for icmp_seq 427
Request timeout for icmp_seq 428
64 bytes from 52.218.20.116: icmp_seq=429 ttl=34 time=201.169 ms
64 bytes from 52.218.20.116: icmp_seq=430 ttl=34 time=204.454 ms
64 bytes from 52.218.20.116: icmp_seq=431 ttl=40 time=200.767 ms
64 bytes from 52.218.20.116: icmp_seq=432 ttl=40 time=200.806 ms
64 bytes from 52.218.20.116: icmp_seq=433 ttl=40 time=201.198 ms
64 bytes from 52.218.20.116: icmp_seq=434 ttl=40 time=200.990 ms
64 bytes from 52.218.20.116: icmp_seq=435 ttl=40 time=200.958 ms
64 bytes from 52.218.20.116: icmp_seq=436 ttl=40 time=201.723 ms
^C
--- s3.eu-west-1.amazonaws.com ping statistics ---
438 packets transmitted, 429 packets received, 2.1% packet loss
round-trip min/avg/max/stddev = 200.465/284.598/1413.533/238.784 ms

 

[PsyWulf]

Hey @websquadza is there something weird going on internationally?
Keep getting random no route to host responses running a ping to s3 (trying to work on infrastructure hosted in EU)

64 bytes from 52.218.20.116: icmp_seq=425 ttl=34 time=201.349 ms
ping: sendto: No route to host
ping: sendto: No route to host
ping: sendto: No route to host
Request timeout for icmp_seq 427
Request timeout for icmp_seq 428
64 bytes from 52.218.20.116: icmp_seq=429 ttl=34 time=201.169 ms
64 bytes from 52.218.20.116: icmp_seq=430 ttl=34 time=204.454 ms
64 bytes from 52.218.20.116: icmp_seq=431 ttl=40 time=200.767 ms
64 bytes from 52.218.20.116: icmp_seq=432 ttl=40 time=200.806 ms
64 bytes from 52.218.20.116: icmp_seq=433 ttl=40 time=201.198 ms
64 bytes from 52.218.20.116: icmp_seq=434 ttl=40 time=200.990 ms
64 bytes from 52.218.20.116: icmp_seq=435 ttl=40 time=200.958 ms
64 bytes from 52.218.20.116: icmp_seq=436 ttl=40 time=201.723 ms
^C
--- s3.eu-west-1.amazonaws.com ping statistics ---
438 packets transmitted, 429 packets received, 2.1% packet loss
round-trip min/avg/max/stddev = 200.465/284.598/1413.533/238.784 ms

Saw similar to EU,and dns timeouts perplexing

 

[websquadza]

Hey @websquadza is there something weird going on internationally?
Keep getting random no route to host responses running a ping to s3 (trying to work on infrastructure hosted in EU)

64 bytes from 52.218.20.116: icmp_seq=425 ttl=34 time=201.349 ms
ping: sendto: No route to host
ping: sendto: No route to host
ping: sendto: No route to host
Request timeout for icmp_seq 427
Request timeout for icmp_seq 428
64 bytes from 52.218.20.116: icmp_seq=429 ttl=34 time=201.169 ms
64 bytes from 52.218.20.116: icmp_seq=430 ttl=34 time=204.454 ms
64 bytes from 52.218.20.116: icmp_seq=431 ttl=40 time=200.767 ms
64 bytes from 52.218.20.116: icmp_seq=432 ttl=40 time=200.806 ms
64 bytes from 52.218.20.116: icmp_seq=433 ttl=40 time=201.198 ms
64 bytes from 52.218.20.116: icmp_seq=434 ttl=40 time=200.990 ms
64 bytes from 52.218.20.116: icmp_seq=435 ttl=40 time=200.958 ms
64 bytes from 52.218.20.116: icmp_seq=436 ttl=40 time=201.723 ms
^C
--- s3.eu-west-1.amazonaws.com ping statistics ---
438 packets transmitted, 429 packets received, 2.1% packet loss
round-trip min/avg/max/stddev = 200.465/284.598/1413.533/238.784 ms

Looking into this now. MTRs aren't showing any problems here now. This hits AWS at peering level, so may have been offnet. Any other IPs affected?

Saw similar to EU,and dns timeouts perplexing

 

[Chiller89]

Was just using S3 in EU as an example, actually been working on infra in Azure, so can't give specific IPs right now.
Weirdly enough calmed down a few minutes ago and seems alright now.
Will report back if it happens again.

 

[RazorCPT]

Hey @websquadza any news on you guys being available on SADV?

 

[DarkSt0rm]

Hey uhhh, @websquadza, ever got stock of those server CPUs? Think the delivery is pretty late at this point

 

[Deadmanza]

Anyone else having horrendous buffering with twitch?

@websquadza

 

[DarkSt0rm]

Internet is just generally slow at the moment. I'm trying to upload 100GB to the EU at quite literally 50-100KB/s... I'm about to have a mental breakdown...

 

[websquadza]

Internet is just generally slow at the moment. I'm trying to upload 100GB to the EU at quite literally 50-100KB/s... I'm about to have a mental breakdown...

Anyone else having horrendous buffering with twitch?

@websquadza

We're doing a little maintenance to our routers in JHB. This means we're down 2 transit providers here and it meant bringing up a failover transit provider to ensure we're still protected. However, there's a little performance knock there. I've done some basic traffic steering which should assist with upstream bandwidth. We can't adjust advertisements too much in this configuration. We're anticipating another hour of maintenance and everything should be back to normal.

 

[Napalm]

So where's the Multigigs at? I see American and UK people sporting with 10gbit connections on fibre. They also boasting about the 5G speeds which seems to be around 1.6gbit to 1.8gbit.

Surely there must be options of 2gbit in SA already ?

 

[DarkSt0rm]

So where's the Multigigs at? I see American and UK people sporting with 10gbit connections on fibre. They also boasting about the 5G speeds which seems to be around 1.6gbit to 1.8gbit.

Surely there must be options of 2gbit in SA already ?

We don't have '1Gbit ' technically. Its 1Gbit locally and whatever it was internationally. I'd just like symmetrical 1Gbit please and thank you.

 

[MDE]

https://twitter.com/i/web/status/1226779688196100099

 

[Napalm]

omw.. slowness wars continues..

 

[DarkSt0rm]

omw.. slowness wars continues..

aka my slow decent into insanity...

 

[Napalm]

aka my slow decent into insanity...

People don't kill people. Latency kills people.

People getting lag go nutty-fruit-salad.

 

[DarkSt0rm]

People don't kill people. Latency kills people.

People getting lag go nutty-fruit-salad.

I'm still uploading that 100GB file. Split it into 100 100MB files, going to take days. I'm quite literally about to buy a plane ticket and fly out with a USB...

 

[Napalm]

I'm still uploading that 100GB file. Split it into 100 100MB files, going to take days. I'm quite literally about to buy a plane ticket and fly out with a USB...

Tie it to a dove.. like back in the old days.

for real... Better to courier it.