Open hartfordfive opened 7 years ago
Agreed this is a bug. The monitoring API should resolve the variables defined in the config.
This looks like it might be related to #6696
@robbavey looks like #6696 was just an error with the validation, see PR https://github.com/elastic/logstash/pull/7411, I think this is different right?
@robbavey is this resolved?
@jordansissel It isn't - there was some refactoring that meant that the linked fix no longer works. I had a quick look, and it should be simple to rework to fix it up
Hi Elastic Team. Any update on this issue?
Edit: I'm using Logstash 6.3 and bug still seems to exist - can't seem to use any variable for this field, just sends through as %{[fieldname]}.
Edit2: Also tried %{fieldname}, same deal. The exact same variable is used for index and that works fine so I know it's accessible and correct.
Is there a workaround? I've tried to substitute credentials for basic authentication using environment variables in the http input plugin config.
When loading the node pipeline stats via
/_node/stats/pipeline
, environment variables used within the ID field of the plugin don't seem to be interpolated. In my case, I use a env var${LOGSTASH_TENANT}
to indicate the tenant name, when I have multiple internal tenants (in a Docker use-case).Version: 5.2.2 Operating System: Darwin
Sample Logstash Config: (indexing.conf)
Sample Logstash Startup Command (for testing):
Sample Logstash Stats: