Xneelo Outage

[Deckert]

Xneelo has been having a total connectivity outage since 08:40 this morning.

Their status page confirms this:

https://xneelostatus.co.za/incidents/5afc3a47-7809-483a-a1cc-98886971bfd9

But the status page does not mention that co-located servers are also unreachable. Outage is ongoing.

--deckert

 

[Sinbad]

more ddos?

 

[quovadis]

more ddos?

50-90% packet loss on both my uplinks - I'd guess it's likely.

 

[Deckert]

more ddos?

Xneelo isn't saying, but it's definitely starting to look that way.
So much for their DDoS mitigation service.

--deckert

 

[WHG]

Xneelo has been having a total connectivity outage since 08:40 this morning.

Their status page confirms this:

https://xneelostatus.co.za/incidents/5afc3a47-7809-483a-a1cc-98886971bfd9

But the status page does not mention that co-located servers are also unreachable. Outage is ongoing.

--deckert

Must be their whole network, managed servers also impacted.

 

[Deckert]

 

[quovadis]

Must be their whole network, managed servers also impacted.

Definitely is. It's just frustrating that they don't update the advisory to include all services.

 

[MrGray]

 

[Deckert]

What is interesting is that we also have hosted servers at another location and from there to Xneelo, there is zero packet-loss. I guess it's a direct peering link from there to Xneelo.

--deckert

 

[ProtechIT]

Our Colocation servers also affected - let’s see how long this goes on for-

 

[quovadis]

What is interesting is that we also have hosted servers at another location and from there to Xneelo, there is zero packet-loss. I guess it's a direct peering link from there to Xneelo.

--deckert

Definitely protocol and route-dependent. My site-to-site VPN's in and out of Xneelo have no packet loss over the same interfaces so UDP seems unaffected...

 

[Deckert]

Looks like they pulled all of their routes outside of SA. As a result, local connectivity has been fully restored, but international access is completely gone.

 

[Gtx Gaming]

looking a lot more stable on managed server side

 

[Deckert]

As of ~two minutes ago, international connectivity has also been restored.

 

[quovadis]

Yoh, my utilization spiked from 8am to 12pm - About 40GB total during this fiasco.

 

[quovadis]

Here we go again?

 

[TheOracle]

 

[quovadis]

 

[Kosmik]

Yep, luckily redundant connections are still fully active from this morning ourside.

 

[ProtechIT]

So there mitigation isn’t actually working- was resolved for a few hours cause the attack stopped. Awesome stuff