You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It would be better if, for example, "State" could be "applied" and "Response" could contain something like "Environment testing_webhook_upgrade is now at 80ba8d60a3545076534ef551f959a8c551c259f0"
Output into Systemd service logs
With " verbose: false" and with "verbose: true" output is the same:
$ systemctl status puppet-webhook.service
● puppet-webhook.service - Puppet Deployment API Server
Loaded: loaded (/etc/systemd/system/puppet-webhook.service; enabled; vendor preset: enabled)
Active: active (running) since Thu 2023-10-19 07:37:27 UTC; 49s ago
...............................
Oct 19 07:37:27 puppet7 systemd[1]: Started Puppet Deployment API Server.
Oct 19 07:37:27 puppet7 webhook-go[3185508]: time="2023-10-19T07:37:27Z" level=info msg="start queue with 1 jobs"
Oct 19 07:37:27 puppet7 webhook-go[3185508]: time="2023-10-19T07:37:27Z" level=info msg="Worker is waiting for jobs"
Oct 19 07:38:08 puppet7 webhook-go[3185508]: [GIN] 2023/10/19 - 07:38:08 | 202 | 1.402347ms | 104.192.140.246 | POST "/api/v1/r10k/environment"
Oct 19 07:38:08 puppet7 webhook-go[3185508]: time="2023-10-19T07:38:08Z" level=info msg="Worker picked Job 72b78852-6e52-11ee-8cdb-02b1b244607d"
I.e. i can see that first webhook-go sends response to Bitbucket server and only after that Worker picking up "job", but in logs there are no indications if Job succeded or not.
It would be good if in case when webhook-go was able to apply/deploy code -> there would be second entry with clear indication that code was applied/deployed successfully
The text was updated successfully, but these errors were encountered:
@sahaqaa I will look into this. The issues I do see with this request at first glance are:
Changing the State to "applied" when it's added to the queue is not an accurate representation of the State.
I can't speak to BitBucket server, but GitLab and GitHub only allow a single response to a webhook call and that response has to come within 10 seconds. I imagine there's a similar requirement for Bitbucket.
As for the Response field, I will look into what's going on there. It should at least send a response stating what happened with the request. But a full tracking of state would require a separate service to report to.
Hello, we've previously used version "2.0.1", now we are upgrading to "2.2.0". Full current config:
When testing we've seen next output in logs of Bitbucket server:
and in logs of systemd service:
What seems to be not right:
Currently
I believe in case webhook-go was able to process/deploy without errors - it may be better to send different "Response" and different "State".
When i change "verbose: true" response in Bitbucket server logs:
It would be better if, for example, "State" could be "applied" and "Response" could contain something like "Environment testing_webhook_upgrade is now at 80ba8d60a3545076534ef551f959a8c551c259f0"
With " verbose: false" and with "verbose: true" output is the same:
I.e. i can see that first webhook-go sends response to Bitbucket server and only after that Worker picking up "job", but in logs there are no indications if Job succeded or not.
It would be good if in case when webhook-go was able to apply/deploy code -> there would be second entry with clear indication that code was applied/deployed successfully
The text was updated successfully, but these errors were encountered: