Repository navigation
Nginz Pod Error while installing demo wire-server through wire-server-deploy #619
Description
Basic information
- On-premises: yes
- Cloud-Provider:AWS, GCP, none, etc.
- Installation type: demo
- Kubernetes version: <!-- version.Info{Major:"1", Minor:"19", GitVersion:"v1.19.7", GitCommit:"1dd5338295409edcfff11505e7bb246f0d325d15", GitTreeState:"clean", BuildDate:"2021-01-13T13:23:52Z", GoVersion:"go1.15.5", Compiler:"gc", Platform:"linux/amd64"}'
- Helm version: version.BuildInfo{Version:"v3.8.1", GitCommit:"5cb9af4b1b271d11d7a97a71df3ac337dd94ad37", GitTreeState:"", GoVersion:"go1.17.7"}
- Installed with Kubespray: yes
- (Helm) Charts version: 4.35.0'
- List of installed top-level charts: database-ephemeral, demo-smtp
- Other related technologies + version: Ansible, Helm, Terraform
What is the expected result?
nginz pod failing and crashing, in logs i have these things
kubectl logs nginz-cfd66dff6-hjccg
error: a container name must be specified for pod nginz-cfd66dff6-hjccg, choose one of: [nginz-disco nginz]
kubectl logs nginz-cfd66dff6-hjccg --all-containers
Starting nginx
Setting up watches for /etc/wire/nginz/upstreams
nginx PID: 7
Setting up watches.
Watches established.
nginx: [emerg] failed to load keystore [1] in /etc/wire/nginz/conf/nginx.conf:183
/nix/store/b6v3ffbwxi028wdhhrxdmg4ywfj6yngs-reload-script/bin/.nginz_reload.sh-wrapped: line 12: kill: (7) - No such process
Files /etc/wire/nginz/upstreams/upstreams.conf and /etc/wire/nginz/upstreams/upstreams.conf.new differ
upstream change found, replacing /etc/wire/nginz/upstreams/upstreams.conf with /etc/wire/nginz/upstreams/upstreams.conf.new
done, sleeping 2.841718
.
.
.
Setting up watches for /etc/wire/nginz/upstreams
Starting nginx
nginx PID: 7
Setting up watches.
Watches established.
nginx: [emerg] failed to load keystore [1] in /etc/wire/nginz/conf/nginx.conf:183
/nix/store/b6v3ffbwxi028wdhhrxdmg4ywfj6yngs-reload-script/bin/.nginz_reload.sh-wrapped: line 12: kill: (7) - No such process
Activity
[-]Bug: [BUG TITLE][/-][+]Nginz Pod Error while installing demo wire-server through wire-server-deploy[/+]mflib commentedon Jun 19, 2023
When i bash into nginz container i dont see folder /etc/wire/nginz/secrets/zauth.conf, for this reason nginz pod is failing as its not able to find the keystore file. Am i missing any step for demo installation which is causing this issue to happen?
mflib commentedon Jun 19, 2023
Okay now i got this nginz pod to running status. The thing which i tried was to copy zauth.text files public and private keys to brig and nginz section of secrets.yaml file. Then i did helm uninstall wire-server && install wire-server, it got to work. Also i bashed into nginz container to verify and saw it now has those folders and files /etc/wire/nginz/secrets/.
Maybe as docs are changing this step might have got retired.
mflib commentedon Jun 20, 2023
Request URL: https://nginz-https.example.com/v3/properties/webapp
Request Method: GET
Status Code: 404
Websocket connections are failing.
supersven commentedon Jun 20, 2023
Hey @mfarhanmail 👋
*.example.comlooks like a default value that surely does not fit to your setup. You may want to exchange all occurrences ofexample.comwith your domain. 😄mflib commentedon Jun 20, 2023
OK i dint had domains, as of now i have setup DNS record in my local machine in /etc/hosts/ to point to server where wire-server is deployed, so i was able to create and validate user but got stuck in a page where i only see "checking for new messages".
I see other API calls are just working fine, example
Request URL: https://nginz-https.example.com/v3/self
Request Method: GET
Status Code: 200
Request URL: https://nginz-https.example.com/v3/clients/63d004dfd75c4c8d/prekeys
Request Method: GET
Status Code: 200
You mean other features will require correct domain in setup to work? @supersven
Thanks for your guidance.
supersven commentedon Jun 20, 2023
Ah, I see. You may get away with fake domains as long as you stay consistent. However, this is not officially supported. And, you may need to solve HTTPS/SSL certificate issues. (Please, don't do this for a production environment.)
The next debugging step would likely be the open the
webapp(likelyhttps://webapp.example.com) and see which requests are failing.mflib commentedon Jun 21, 2023
Request URL: https://nginz-https.example.com/v3/properties/webapp
Request Method: GET
Status Code: 404
Request URL: https://nginz-https.example.com/v3/properties/WIRE_RECEIPT_MODE
Request Method: GET
Status Code: 404
Request URL: https://nginz-https.example.com/v3/properties/WIRE_TYPING_INDICATOR_MODE
Request Method: GET
Status Code: 404
Websocket:
An SSL certificate error occurred when fetching the script.
reconnecting-websocket-mjs.js:518 WebSocket connection to 'wss://nginz-ssl.example.com/await?access_token=0fIw2u2Ewm-C5RgV9o30lCs4o1TpAXAEaBtor2DqSL4SkKnXLvfz3kaWqCw9hBf81XvzvSsa21lemPBC_rZFDg==.v=1.k=1.d=1687319583.t=a.l=.u=0e1a7141-4a46-4201-abf4-324a25cab575.i=63d004dfd75c4c8d.c=13529975733778401668&client=63d004dfd75c4c8d' failed:
Rest of the api's are working fine with 200 status code in the page where still "Checking for new messages" progress is shown.
@supersven
julialongtin commentedon Jun 21, 2023
what have you done for certificates? if you're doing the "just hit the trust button" method, you need to go to nginz-ssl.example.com manually in a web browser, and hit trust there.
mflib commentedon Jun 21, 2023
Thank you it worked amazingly. Now ill get away with fake certificates.