aws-samples / aws-xray-kubernetes Goto Github PK

View Code? Open in Web Editor NEW

36.0 36.0 42.0 90 KB

Code examples showing how to run AWS X-Ray on a Kubernetes cluster for deep application insights.

Home Page: https://aws.amazon.com/blogs/compute/application-tracing-on-kubernetes-with-aws-x-ray/

License: Apache License 2.0

JavaScript 70.95% Dockerfile 29.05%

aws aws-xray kubernetes

aws-xray-kubernetes's People

Contributors

Stargazers

Watchers

Forkers

jicowan getamazednow andrewd-uriux ssro tuapuikia yutachaos larssonandreas kanwaljitkhurmi samuraisalad guillaumesmaha pksinghus gmagella-ca jorgebaruchi abulkaware n0n0x the-codefactory-dev collaborne evertonteotonio anxious-coder-lhs umeshkumar04 mikeapted ramkarthik-sys bablubharath bacteria673 aa-tt jalawala yuriiholiuk ckassen akramdhiabi sharplee7 tetsuo-nobe 7sonny7 ved-kickdrum damyan-h lzjona mhndsingh21 ribeiro

aws-xray-kubernetes's Issues

XRay Daemon routing issue

I tired to reproduce what was presented in this blog post and at the first sight all seems to look ok.

But after verification, I discovered, that when serviceA is accessing xray daemon it is doing it via kubernetes service - so it is loadballanced, and request can be routed to daemon deployed on other node than the application.

So instead of this:

It looks more like this:

When deploying the container using a DaemonSet, the X-Ray port is also exposed directly on the host. This way, clients can connect directly to the daemon on your node. This avoids unnecessary network traffic going across your cluster.

So my question is how to achieve this? How to connect to xray on same node.

Missing serviceAccountName for deployment if using service account to grant xray permission

The xray-daemon yaml misses the serviceAccountName if using service account.

  template:
    metadata:
      labels:
        app: xray-daemon
    spec:
     # add below line if using service account to grant xray permissions to pod
      serviceAccountName: xray-daemon 
      volumes:
        - name: config-volume
          configMap:
            name: "xray-config"
      containers:

Communication issue with XRay Daemon

Following the instructions in the repo and the blog post results in some issues. Not sure if it's only in my environment.

The services (service-a and service-b) don't start. See below.

NAMESPACE     NAME                             READY   STATUS             RESTARTS   AGE
default       pod/service-a-8699d4bf68-65f5x   0/1     CrashLoopBackOff   11         32m
default       pod/service-a-8699d4bf68-hh2l5   0/1     CrashLoopBackOff   11         32m
default       pod/service-b-86564fd46b-b4ft8   0/1     CrashLoopBackOff   11         32m
default       pod/service-b-86564fd46b-kvplc   0/1     CrashLoopBackOff   11         32m

demo-app ➤ kubectl logs  pod/service-a-8699d4bf68-65f5x
/usr/src/app/node_modules/aws-xray-sdk-core/lib/daemon_config.js:26
    throw new Error('Invalid Daemon Address. You must specify an ip and port.');
    ^

Error: Invalid Daemon Address. You must specify an ip and port.
    at processAddress (/usr/src/app/node_modules/aws-xray-sdk-core/lib/daemon_config.js:26:11)
    at Object.<anonymous> (/usr/src/app/node_modules/aws-xray-sdk-core/lib/daemon_config.js:62:3)
    at Module._compile (module.js:653:30)
    at Object.Module._extensions..js (module.js:664:10)
    at Module.load (module.js:566:32)
    at tryModuleLoad (module.js:506:12)
    at Function.Module._load (module.js:498:3)
    at Module.require (module.js:597:17)
    at require (internal/module.js:11:18)
    at Object.<anonymous> (/usr/src/app/node_modules/aws-xray-sdk-core/lib/segment_emitter.js:105:17)

There seems to be an issue communicating with the X-Ray daemonset. However, the it seems to be running properly.

demo-app ➤ kubectl logs pod/xray-daemon-5v7zt
2020-03-27T14:53:29Z [Info] Initializing AWS X-Ray daemon 2.1.3
2020-03-27T14:53:29Z [Info] Using buffer memory limit of 76 MB
2020-03-27T14:53:29Z [Info] 1216 segment buffers allocated
2020-03-27T14:53:29Z [Info] Using region: us-east-2

Another recommendation I have is to update the K8s deployment template to v1 instead of v1beta so we can use selectors

Unsure of need for cluster-admin / host networking

From looking at documentation, it doesn't look like the xray daemon needs cluster-admin permissions (or any kubernetes api access whatsoever) or host networking enabled. Is there anything I've missed or can those be safely removed from the manifest?

aws-samples / aws-xray-kubernetes Goto Github PK

aws-xray-kubernetes's People

Contributors

Stargazers

Watchers

Forkers

aws-xray-kubernetes's Issues

XRay Daemon routing issue

Missing serviceAccountName for deployment if using service account to grant xray permission

Communication issue with XRay Daemon

Following the instructions in the repo and the blog post results in some issues. Not sure if it's only in my environment.

Another recommendation I have is to update the K8s deployment template to v1 instead of v1beta so we can use selectors

Unsure of need for cluster-admin / host networking

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent