Missing labels

[mscdex]

I noticed that in some cases, like this PR, the bot failed to label the PR at all. It should have added the test label.

[sejoker]

I am curious how to check log files in case we got an error from Github. In that case, we should have a message 'Error retrieving files from GitHub' in the logs or any other relevant debug information about the PR in question.

[mscdex]

Your guess is as good as mine. I've never tried to run my own instance of the bot before.

[jbergstroem]

We use Bunyan so logs are available on the host as well as over http (with a secret). I'm sure @phillipj can provide the access key to those that wants to fiddle.

[phillipj]

Absolutely! Just ping me on IRC and I'll give you the URL

On Sunday, 18 September 2016, Johan Bergström notifications@github.com
wrote:

We use Bunyan so logs are available on the host as well as over http (with
a secret). I'm sure @phillipj https://github.com/phillipj can provide
the access key to those that wants to fiddle.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#79 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ABLLExztPLaj9W8Nr2ysYZLLzFhGkroHks5qrVasgaJpZM4J_nO0
.

[phillipj]

Had a look in the logs and found this:

[2016-09-17T11:42:06.032+02:00] ERROR: bot/15719 on server-05: Error retrieving files from GitHub (req_id=8dfe80e7-e6dd-4144-9eff-5c81afb17e88, pr=nodejs/node/#8600, action=pull_request.opened)
    err: {
      "code": 404,
      "status": "Not Found",
      "message": "{\"message\":\"Not Found\",\"documentation_url\":\"https://developer.github.com/v3\"}"
    }

I assume there was some inconsistencies in GitHub's side which caused this error, while trying to retrieve the files related to the mentioned PR :/

Maybe the bot should wait a couple of seconds before doing its thing, as it's not important that these labels gets added instantaneously.

[sejoker]

@phillipj thanks for follow up, you confirmed my initial assumption. A short timeout like you suggested would be sufficient as the first step. If necessary, more complicated setup with retry logic will need to take place.

[sejoker]

I could prepare a PR if we agree on the possible fix.

[phillipj]

@sejoker a PR would be great 👍

[sejoker]

@phillipj your feedback on #81 will be valuable.