This works as is, but I recommend setting the LETS_ENCRYPT_ENV=staging as the resolver has not been actively disabled here!
Traefik is at v2.10.7 for my testing, wasn't sure how to format the version and my brain is barely working now...
- This works as is, but I recommend setting the `LETS_ENCRYPT_ENV=staging` as the resolver has not been actively disabled here!
- Traefik is at v2.10.7 for my testing, wasn't sure how to format the version and my brain is barely working now...
If I understand you correctly, you are asking to have two version vars? I already included SECRET_WILDCARD_CERT_VERSION, used for both as typically both change for me at the same time.
I guess in theory you could reuse the pwd/key, is that correct, or is it just "best practice" to allow a version var per variable, regardless of anything else?
If I understand you correctly, you are asking to have two version vars? I already included SECRET_WILDCARD_CERT_VERSION, used for both as typically both change for me at the same time.
I guess in theory you could reuse the pwd/key, is that correct, or is it just "best practice" to allow a version var per variable, regardless of anything else?
Generally it's env var per secret, yep. So more convention at this point. But if you feel strongly against, please then document and sure just move on 👍
Generally it's env var per secret, yep. So more convention at this point. But if you feel strongly against, please then document and sure just move on 👍
is the correct method to use/all that is needed. I'm pretty new to traefik still however, so ...🧂
My understanding is that the already existing:
```yaml
web:
address: ":80"
http:
redirections:
entryPoint:
to: web-secure
```
is the correct method to use/all that is needed. I'm pretty new to traefik still however, so ...🧂
If you were asking me: yes, I think it's good. Always nice to get a thumbs up from CI, but I'm guessing that's currently broken (or maybe won't load vars against a PR)?
If you were asking me: yes, I think it's good. Always nice to get a thumbs up from CI, but I'm guessing that's currently broken (or maybe won't load vars against a PR)?
For "security" reasons, otherwise PRs (which could come from anyone, including people who log in via Gitlab / Github) could exfiltrate secrets, including SSH keys for the test server, and login credentials to publish Gitea packages.
> or maybe won't load vars against a PR
Indeed it's this ☝️
For "security" reasons, otherwise PRs (which could come from anyone, including people who log in via Gitlab / Github) could exfiltrate secrets, including SSH keys for the test server, and login credentials to publish Gitea packages.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
LETS_ENCRYPT_ENV=stagingas the resolver has not been actively disabled here!Looking good, tysm for working on this!
@@ -0,0 +9,4 @@secrets:ssl_cert:name: ${STACK_NAME}_ssl_cert_${SECRET_WILDCARD_CERT_VERSION}${STACK_NAME}_ssl_cert_${SECRET_SSL_CERT_VERSION}I think you need to make a specific env var for this in the
.env.samplealso?If I understand you correctly, you are asking to have two version vars? I already included SECRET_WILDCARD_CERT_VERSION, used for both as typically both change for me at the same time.
I guess in theory you could reuse the pwd/key, is that correct, or is it just "best practice" to allow a version var per variable, regardless of anything else?
Generally it's env var per secret, yep. So more convention at this point. But if you feel strongly against, please then document and sure just move on 👍
@@ -0,0 +12,4 @@name: ${STACK_NAME}_ssl_cert_${SECRET_WILDCARD_CERT_VERSION}external: truessl_key:name: ${STACK_NAME}_ssl_key_${SECRET_WILDCARD_CERT_VERSION}name: ${STACK_NAME}_ssl_key_${SECRET_SSL_KEY_VERSION}I think you need to make a specific env var for this in the
.env.samplealso?@@ -25,7 +25,6 @@ http:security:headers:frameDeny: truesslRedirect: trueDoes this break some expectations?
My understanding is that the already existing:
is the correct method to use/all that is needed. I'm pretty new to traefik still however, so ...🧂
For reference, this is the message I was getting:
level=warning msg="SSLRedirect is deprecated, please use entrypoint redirection instead."Great stuff, let's drop that and hope 🏄♀️
5e7e16a707toe3c1df83faLGTM, looks safe to merge, shall we? Thanks!
If you were asking me: yes, I think it's good. Always nice to get a thumbs up from CI, but I'm guessing that's currently broken (or maybe won't load vars against a PR)?
I've checked on my end too your changes @wolcen and they look good 👍
Indeed it's this ☝️
For "security" reasons, otherwise PRs (which could come from anyone, including people who log in via Gitlab / Github) could exfiltrate secrets, including SSH keys for the test server, and login credentials to publish Gitea packages.