Downloading a large file seems slow ?

[Izopi4a]

First: Awesome work guys. I love it ! However i am experiencing a small issue

Is this a BUG REPORT or FEATURE REQUEST? (choose one):

I am not sure

NGINX Ingress controller version:
0.20.0

Kubernetes version (use kubectl version):

Client Version: version.Info{Major:"1", Minor:"12", GitVersion:"v1.12.2", GitCommit:"17c77c7898218073f14c8d573582e8d2313dc740", GitTreeState:"clean", BuildDate:"2018-10-24T06:54:59Z", GoVersion:"go1.10.4", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"12", GitVersion:"v1.12.2", GitCommit:"17c77c7898218073f14c8d573582e8d2313dc740", GitTreeState:"clean", BuildDate:"2018-10-24T06:43:59Z", GoVersion:"go1.10.4", Compiler:"gc", Platform:"linux/amd64"}

Environment:

• Cloud provider or hardware configuration: hardware machine
• OS (e.g. from /etc/os-release): ubuntu 18.04
• Kernel (e.g. uname -a): 4.15.0-38-generic Admin documentation #41-Ubuntu SMP Wed Oct 10 10:59:38 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
• Install tools: kubeadm
• Others:

I have a server ( core 2 duo with 8 G ram / i know it is low but it seems enough for my purposes/). We are in LAN with the server with 1 GB/s connection.
I used this command in my pod to generate a large file in my running pod

dd if=/dev/zero of=outputFile bs=2G count=1

than after opening the correct url ( https is enabled ) in my browser, a download happens. What I get is download speed of about 6-8 MB/s. Now I replaced nginx ingress controller with traefik and set the resource limit to CPU: 800m / 1000Mi RAM, and i get ~ 60 MB/s. I did try to play with resources for nginx ingress but still the same result.

Here is my POD nginx configuration ( nothing special in the server config )

user www-data;
pid /run/nginx.pid;
worker_processes 1;
events {
    worker_connections 1024;
}
http {
    sendfile on;
    tcp_nopush on;
    tcp_nodelay on;
    keepalive_timeout 65;
    types_hash_max_size 2048;
    server_tokens off;
    include /etc/nginx/mime.types;
    default_type application/octet-stream;
    ...

I used helm to install nginx ingress here is the command

helm install stable/nginx-ingress --name nginx-ingress --namespace kube-system -f nginx_values.yaml
#######
cat nginx_values.yaml

controller:
  kind: "DaemonSet"
  daemonset:
    useHostPort: true
  service:
    type: "NodePort"

So everything is identical in both cases /just swapping ingresses/ , but I get much lower download speed with nginx ingress. Am I doing something wrong, or this is how it is supposed to work ?

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[msaspence]

@Izopi4a did you have any luck with this in the end?

[Izopi4a]

@msaspence nope, but I didn't try tbh. Both test were identical - yet the difference so big, so I start using traefik, as this test was suspicious.

//p.s. i think its just cuz traefik is multi-threading a lot. I haven't written down the speeds, but i think nginx did like 5-6 mb/s while traefik 20+ with same cpu req/limit