ansible: on the first run fluentd cannot reload properly

[xlr-8]

For some currently unknown reason, fluentd cannot reload on the first run, but re-runing ansible passes just fine.

RUNNING HANDLER [cycloid.fluentd : reload td-agent] ****************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
Monday 03 September 2018 09:46:58 +0000 (0:00:00.033) 0:05:19.906 ******
fatal: [10.1.1.24]: FAILED! => {"changed": false, "msg": "Unable to reload service td-agent: Job for td-agent.service failed. See 'systemctl status td-agent.service' and 'journalctl -xn' for details.\n"}

[Seraf]

It should have been fixed with the new releases of fluentd, there was a bug on the aws plugin

treasure-data/omnibus-td-agent#184

Let's reopen this issue if the problem raise again.