Extremely high web server request time


Since about 9:00 AM EST this morning, my NGINX web server located in Paris has been seeing very high request times (2 - 50 seconds) for uncached traffic coming from the US, mostly from bots (Google, Bing, Pinterest, etc.).

I contacted my host to check the status of the server, and everything is working perfectly. The total load is very low (CPU around 0.5%), and the logs are clean and show no problems. The server configuration hasn’t changed, and it had very little traffic today, so high traffic is clearly not the culprit.

The reason I’m coming here is because I see there’s an open Cloudflare issue:

Network Performance Issues

Investigating - Cloudflare is investigating issues with network performance. Customers may see increased 5xx errors and/or latency.

I’m wondering if this might have anything to do with the problem I’m facing?

Cached pages coming from Cloudflare are getting a low request time (< 0.40 ms), so I’m thinking the issue might be between Cloudflare and my origin. It could also be a network issue between the United States and Paris, but all the pings and traceroutes I’ve performed seem to be perfectly fine (~100ms latency).

Any suggestions to other tests I could perform to find out if the problem lies between Cloudflare and my origin would be greatly appreciated.


I’ve just performed an MTR test between my web server and my Cloudflare IP, and I can see that there’s huge packet loss in France:

Packet Loss %
cloudflare.par.franceix.net 41.8%

So, if anyone else in France is having similar issues, there is currently important packet loss on one of Cloudflare’s servers.


I’ve opened a ticket with Cloudflare in case this issue is unrelated to the current ongoing problem.


I now see that the traffic has been rerouted to cloudflare-2.par.franceix.net and there is 0% packet loss.

I will monitor the situation on my end and provide further comments if needed.


Is there an attack in progress?

I now see 10-1-7.ear2.Paris1.Level3.net with a packet loss of 26.8%. This is a server located in the route between bots from the United States and my server in Paris. This explains the high request time they are experiencing.


Three other servers with 40% packet loss: ae-7.r21.parsfr04.fr.bb.gin.ntt.net, ae-13.r24.asbnva02.us.bb.gin.ntt.net and prs-bb2-link.ip.twelve99.net


The outage seems to be affecting different parts of Europe and not just France. London server ldn-bb2-link.ip.twelve99.net has 92% packet loss.

Anyone knows what’s going on? Or where to report these outages?

1 Like

Hi there,

I do suspect the incident may have some responsibility for your trouble.
Please do check up on the status page and try again after the incident calms down.

Thank you.

1 Like