awx job/host monitoring details

HI there,

I have a few AWX jobs that run on some pretty sluggish hosts (legacy stuff), unfortunately I have a few hundred hosts. Currently I run with serial set to 20 or so which works well and fine.

The issue I have is monitoring executions for critical deployments as it can take an hour or so to run. It would be nice to have a cli command or gui visibility where I can ask it to print out the current status of all the hosts so I can easily tell that say 50% of the hosts are done, 30 failed etc. In AWX this information is only available after the playbook completes, i’m looking to get this information while the job template is running.

ie… something like:
tower-cli job status --hostsummary 1833

==== ============ =========================== ======= ==========
host . batch . created status elapsed
==== ============ =========================== ======= ==========
1833 1 host1 2019-01-18T22:06:31.654397Z success 1586.83164
1833 1 host2 2019-01-18T22:06:31.654397Z failed 1586.83164

1833 1 host3 2019-01-18T22:06:31.654397Z running 1586.83164

1833 2 host4 2019-01-18T22:06:31.654397Z running 1586.83164

1833 2 host5 2019-01-18T22:06:31.654397Z waiting 1586.83164

==== ============ =========================== ======= ==========

Is this something you would see potentially adding into AWX? Does AWX store this information in it’s database? Does it only store it after the whole playbook completes? Is another option to query the db directly? Can this information be aquired currently from the API?

I think I may have a look at ara to see if this might be a complementing technology for this requirement. If anyone also has experience with this would be great to let me know your feedback on how it went.

What I am doing in the meantime is just will use the output of “tower-cli job stdout 1833” and use grep/etc to get the information i need. But if something better was available, that would be great.

The summaries are stored after the job is complete but you could write something that queries the api endpoints for the job events themselves /api/v2/jobs/n/job_events/ will give you what you are looking for.