rud Posts - ZITADEL

rud

•Created by rud on 6/2/2025 in #questions-help-bugs

Only getting 404 from /v2beta/projects/search in v3.2.2?

Following the docs at https://zitadel.com/docs/apis/resources/project_service_v2/project-service-list-projects The query is simply a POST to /v2beta/projects/search with a Bearer token. Grabbed from the docs. What am I missing?

 ruby
BEARER_TOKEN = "QYxS....laRik"
CUSTOM_DOMAIN = "localhost:8080"

require "uri"
require "json"
require "net/http"

url = URI("http://#{CUSTOM_DOMAIN}/v2beta/projects/search")

http = Net::HTTP.new(url.host, url.port);
request = Net::HTTP::Post.new(url)
request["Content-Type"] = "application/json"
request["Accept"] = "application/json"
request["Authorization"] = "Bearer #{BEARER_TOKEN}"
request.body = JSON.dump({})

response = http.request(request)
puts response.read_body
=> {"code":5,"message":"Not Found"}
puts response.code
=> 404

 ruby
BEARER_TOKEN = "QYxS....laRik"
CUSTOM_DOMAIN = "localhost:8080"

require "uri"
require "json"
require "net/http"

url = URI("http://#{CUSTOM_DOMAIN}/v2beta/projects/search")

http = Net::HTTP.new(url.host, url.port);
request = Net::HTTP::Post.new(url)
request["Content-Type"] = "application/json"
request["Accept"] = "application/json"
request["Authorization"] = "Bearer #{BEARER_TOKEN}"
request.body = JSON.dump({})

response = http.request(request)
puts response.read_body
=> {"code":5,"message":"Not Found"}
puts response.code
=> 404

9 replies

ZZITADEL

•Created by rud on 3/7/2025 in #questions-help-bugs

Organization metadata sometimes changed when reading it back

I think this might be a bug in release 2.71.1. I write a value of "5566778899" to an organization metadata field. When I read it back I receive the value "556677889w==". I've observed values roundtrip correctly (shorter values, longer values), but something about this value triggers an edge case. If it matters: I'm accessing the REST API from ruby, Zitadel is running locally in Docker Compose. Sample API interaction attached. Hope this is easy to figure out with a test-case.

5 replies

ZZITADEL

•Created by rud on 9/25/2024 in #questions-help-bugs

Translations - they just keep coming..

Managing translations in a steadily growing list languages seems like it must be somewhat painful, keeping every piece of text in every language up to date, and any change that affects end-user text becomes a massive coordination task. So, I've seen other open source software use weblate.org (you can self-host or pay for their cloud) as a way to fix the continuous translation problem, and even crowd-source translations without people having to know git. I've no idea if this is a good fit for the current workflow for preparing Zitadel releases, but the pain seems real, and I thought I'd share a solution I've seen others benefit from. Cheers!

5 replies

ZZITADEL

•Created by rud on 9/6/2024 in #questions-help-bugs

no pg_hba.conf entry for host - `init`+`setup` works fine with PG, but main `start` entrypoint fails

I'm seeing PostgreSQL connection errors from the zitadel "start" process, and I'm finding them puzzling, as the init and setup processes complete successfully. This is using version ghcr.io/zitadel/zitadel:v2.61.0. What does work: init and setup processes run to completion and can all connect just fine to PostgreSQL and manage to setup the full structure of schemas and tables. What does not work: The main start process boots fine, but shortly after starts emitting messages of this format (IPs anonymized).

time="2024-09-06T11:26:02Z" level=info msg="query failed" caller="/home/runner/work/zitadel/zitadel/internal/eventstore/repository/sql/query.go:129" error="failed to connect to `user=zitadel database=zitadel`: 10.10.10.5:5432 (example.com): server error: FATAL: no pg_hba.conf entry for host \"10.10.10.3\", user \"zitadel\", database \"zitadel\", SSL encryption (SQLSTATE 28000)"

time="2024-09-06T11:26:02Z" level=info msg="query failed" caller="/home/runner/work/zitadel/zitadel/internal/eventstore/repository/sql/query.go:129" error="failed to connect to `user=zitadel database=zitadel`: 10.10.10.5:5432 (example.com): server error: FATAL: no pg_hba.conf entry for host \"10.10.10.3\", user \"zitadel\", database \"zitadel\", SSL encryption (SQLSTATE 28000)"

* Zitadel server: 10.10.10.3 * PostgreSQL server: 10.10.10.5. Version 16. Line from pg_hba.conf which I expected to allow access with the correct password:

# method database user    address       auth-method
hostssl  zitadel  zitadel 10.10.10.0/24 scram-sha-256

# method database user    address       auth-method
hostssl  zitadel  zitadel 10.10.10.0/24 scram-sha-256

I verified the zitadel user in PostgreSQL does indeed use SCRAM-SHA-256:

zitadel=# select rolname,rolpassword from pg_authid where rolname = 'zitadel';
-[ RECORD 1 ]--------------------
rolname     | zitadel
rolpassword | SCRAM-SHA-256$4096:<snipped>

zitadel=# show password_encryption;
-[ RECORD 1 ]-------+--------------
password_encryption | scram-sha-256

zitadel=# select rolname,rolpassword from pg_authid where rolname = 'zitadel';
-[ RECORD 1 ]--------------------
rolname     | zitadel
rolpassword | SCRAM-SHA-256$4096:<snipped>

zitadel=# show password_encryption;
-[ RECORD 1 ]-------+--------------
password_encryption | scram-sha-256

I have served up the connection credentials (stored encrypted) as a snippet of YAML looking like this:

database:
  postgres:
    Host: 10.10.10.5
    Port: 5432
    Database: zitadel
    User:
      Username: zitadel
      Password: TheActualPasswordInPlainTextHere
      SSL:
        Mode: require

database:
  postgres:
    Host: 10.10.10.5
    Port: 5432
    Database: zitadel
    User:
      Username: zitadel
      Password: TheActualPasswordInPlainTextHere
      SSL:
        Mode: require

What am I not seeing? As may be apparent from the above I've been looking at this for a while now 🙂

4 replies

ZZITADEL

•Created by rud on 4/8/2024 in #questions-help-bugs

Version bump k8s helm chart Github Action - do you take patches?

21 replies