reactioncommerce/deployment

The place for Reaction deployment discussions.

kamaljaiswal

@kamaljaiswal

tried hard reset the /home/reaction ..will update .

any help ..as same ansible-playbook playbooks/reaction.yml -l reaction.server was working . And tried deleting /home/reaction completely

kamaljaiswal

@kamaljaiswal

systemctl stop mongodb did the trick ...will confirm in case

seday72

@seday72

hi everyone

i am going to deploy reaction commerce platform on aws ec2

I tried to configure .env but backend not working

please help me how to config this

Pulkit Garg

@wittycodes

@seday72 https://github.com/reactioncommerce/proxy-traefik
refer the above git repo, it's for an automated & workable production deployment of reaction.

seday72

@seday72

This message was deleted

hi @wittycodes

is that possible to deploy on aws too?

Loan Laux

@loan-laux

Hey @seday72, we’ve got a tutorial for deploying on AWS ECS on https://university.outgrow.io

ihalton

@ihalton

@ihalton
Running command: meteor run --port 4080 --no-lint --no-release-check --raw-logs
[[[[[ /usr/local/src/app ]]]]]

=> Started proxy.
Killed
Error: Command failed: REACTION_METEOR_APP_COMMAND_START_TIME=1597467210450 meteor run --port 4080 --no-lint --no-release-check --raw-logs
at checkExecSyncError (child_process.js:630:11)
at Object.execSync (child_process.js:666:15)
at run (file:///usr/local/src/app/.reaction/scripts/run.mjs:21:18)
at file:///usr/local/src/app/.reaction/scripts/run.mjs:29:1
at ModuleJob.run (internal/modules/esm/module_job.js:110:37)
at async Loader.import (internal/modules/esm/loader.js:164:24) {
status: 137,
signal: null,
output: [ null, null, null ],
pid: 39,
stdout: null,
stderr: null
}

Error: App failed to start
hello, get crashed when running reaction_admin from code. why?

balibebas

@balibebas

Hi folks. Has anyone looked at deploying with Rio? https://rio.io

I'd like to take advantage of the canary deployments and GitOps. Rio is a MicroPaaS by Rancher and is designed to work well with k3s. I've done some initial research here for anyone interested who hasn't worked with Rio yet: https://habd.as/post/git-based-continuous-delivery-rio/

ihalton

@ihalton

Hello, have anyone know what reason about the error? Thanks

Running command: meteor run --port 4080 --no-lint --no-release-check --raw-logs
[[[[[ /usr/local/src/app ]]]]]
=> Started proxy.
Killed
Error: Command failed: REACTION_METEOR_APP_COMMAND_START_TIME=1597467210450 meteor run --port 4080 --no-lint --no-release-check --raw-logs
at checkExecSyncError (child_process.js:630:11)
at Object.execSync (child_process.js:666:15)
at run (file:///usr/local/src/app/.reaction/scripts/run.mjs:21:18)
at file:///usr/local/src/app/.reaction/scripts/run.mjs:29:1
at ModuleJob.run (internal/modules/esm/module_job.js:110:37)
at async Loader.import (internal/modules/esm/loader.js:164:24) {
status: 137,
signal: null,
output: [ null, null, null ],
pid: 39,
stdout: null,
stderr: null
}
Error: App failed to start

balibebas

@balibebas

@ihalton I see you asked the same question in the main Reaction channel just an hour before here. Could you please try waiting there for an answer instead of double-posting? My understanding is this channel is not intended for general debugging.

ihalton

@ihalton

@balibebas I am sorry for any inconvenience, I will waiting answer in that channel.

balibebas

@balibebas

thanks, bos. you may wish to try the forums too for visibility

ihalton

@ihalton

@balibebas this maybe it is a bug. Nobody answer me in that channel. :)

Ankur Pata

@ankurpata

Hi @loan-laux
The Reaction-Hydra repo says ' its 💥 Not safe for production.'
What does this exactly mean? Do we need to use our custom authentication service instead of Hydra for production or we need to customize based on our needs ?

balibebas

@balibebas

Food for thought... Coinbase: Why Kubernetes is not part of our stack https://blog.coinbase.com/container-technologies-at-coinbase-d4ae118dcb6c

S Tunji Turner

@stunjiturner

Reaction + Vercel anybody?

Loan Laux

@loan-laux

Yep. Works. I can confirm.

Q_Q

@mfaster

Do anyone still works on Heroku?

I'd like to take advantage of the canary deployments and GitOps. Rio is a MicroPaaS by Rancher and is designed to work well with k3s. I've done some initial research here for anyone interested who hasn't worked with Rio yet: https://habd.as/post/git-based-continuous-delivery-rio/

Good one tho

Q_Q

@mfaster

Thanks for sharing

Karim Abuzaid

@Vercryger

Hey everyone! I've switched to kubernetes for both production and staging enviroments and currently working on setup SSL, which I've found it so hard btw. So far I've secured all services but within the admin the socket connection fails since I changed ws to wss. At first I thought some configurations were wrong on the backend but, storefront successfully connects to the web socket server. I'm not really sure what's going on, this is the error I get failed: Error during WebSocket handshake: Unexpected response code: 400. Any help would be much appreciated

Also, identity server also get this error, so probably meteor related?

Sebastian Franco

@jhonsfran1165

Yeah, It seems some configuration is missing. What API gateway are you using?

Karim Abuzaid

@Vercryger

I'm using EKS and this started to happen when I configured ELB to make use of ssl certificates

For the record I'm not using ingress controllers for such setup

Sebastian Franco

@jhonsfran1165

Okay I can help with this but I need some details about your infra, I also need logs of your services. Did you configure a ingress? Could you share the yaml file? What I do in that situation is configure a node port directly to the service that I need to test and then test the connection. If everything goes well it means that the problems is in the ingress configuration. So you can check that part specifically.

igor romanovsky

@igorromanovsky_twitter

Hey. Is there any guide for deploying to GCP

Karim Abuzaid

@Vercryger

@jhonsfran1165 Thank you. This is the current yaml for our reaction admin, the rest of services are just more of the same

apiVersion: v1
kind: Service
metadata:
  labels:
    reaction: admin
  name: reaction-admin
  annotations:
    service.beta.kubernetes.io/aws-load-balancer-backend-protocol: "http"
    service.beta.kubernetes.io/aws-load-balancer-ssl-cert: arn:aws:acm:us-east-2:<accountno>:certificate/<cid>
    service.beta.kubernetes.io/aws-load-balancer-ssl-ports: "https"
    service.beta.kubernetes.io/aws-load-balancer-proxy-protocol: "*"
spec:
  ports:
  - name: http
    port: 80
    targetPort: 4080
  - name: https
    port: 443
    targetPort: 4080
  selector:
    reaction: admin
  type: LoadBalancer
status:
  loadBalancer: {}

4 replies

Like I mentioned before, we're not using ingress controllers of any kind, we've achieved SSL for all services using just ELB

igor romanovsky

@igorromanovsky_twitter

Hey all...
I'm trying to deploy reaction-admin, reaction and reaction-hydra to GCP. I've deployed reaction-admin and reaction, but now I'm struggling with the deployment of the reaction-hydra.
unfortunately I don't have experience with "hydra" so if someone can give me a smal direction to how it could be done, I would be very grateful.

8 replies

igor romanovsky

@igorromanovsky_twitter

Hey guys. Just wanted to share a really good installation guide on K8s that uses Helm chart :
https://slingshotlabs.io/blog/deploying-reaction-commerce-on-kubernetes/

Andrei Balmus

@abalmush

@igorromanovsky_twitter Hi Igor, thanks for great article. I have a question for the community. Do you know if I can find information on how to configure a continues deployment. For example I would like to change storefront a little bit. What would be the best way for deploying my changes into existing cluster? Sorry I'm new in this and I'm trying to understand please help!

3 replies

Andrei Balmus

@abalmush

Ok I think I understand I have to generate new version of docker image and push it to docker registry and then use it

Andrei Balmus

@abalmush

Hi guys, I successfully deployed reaction to kubernetes cluster and everything works as expected. BUT if I use existing mongo db instead of local one, I'm getting this error.

The db user has admin privileges, what I'm doing wrong please help.

MitchellWT

@MitchellWT

Hey, I'm attempting to deploy an online development platform using digital ocean. I'm using this github repo https://github.com/reactioncommerce/proxy-traefik. The installation works fine until I perform a "make init-dev", after that point all web addresses (except the traefik dashboard) respond with traefik's "404 page not found" page. I think the issue is that traefik can't locate the new files. So I'm wondering where I can find where traefik determines the file locations for web addresses?

l2ysho

@l2ysho

The db user has admin privileges, what I'm doing wrong please help.

Hi, you need also clusterAdmin and dbAdmin roles for your user

Robin T

@qispark

Hi there, I was wondering if anyone encountered an issue with Reaction Admin using a production Reaction server. In my docker instance, when I point it to my Reaction server running on localhost in docker, it works. However, it seems like it's unable to generate a valid token when configured with a remote production Reaction server, causing the user to log out after having just logged in. Just wondering if anyone has encountered such an issue and what sort of misconfiguration may cause this?

2 replies

Where communities thrive

People

Repo info

Activity