Tool to make eck diag a bit more human readable

git clone https://github.com/jlim0930/eck-glance.git

Goto the extracted eck-diag and run /path/eck_1.sh for most cases.

If your workstation is able to handle it you can run /path/eck_1fast.sh it will launch all the jobs in the background and run all the subscripts at once - I tried this on multiple diags and it was super fast !!

All eck_*_1.sh scripts will give high level of the kind that you are looking at(more detailed than just kubectl get <kind>) All eck_*_2.sh will give a more descriptive and detailed view of the resource(its more like kubectl describe <kind> <resource>)

Files generated by the tool will be named eck_*.txt

In the main ECK diag directory eck_nodes.txt

In each name space any elastic related items will be eck_[elasticsearc|beat|agent|kibana|enterprisesearch/elasticmapservice]*.txt All other eck_*.txt are for kubernetes kinds

• StatefulSets - manages elasticsearch
• ReplicaSets - manages beats(metricbeats/etc that runs on 1 or more host)/apmserver/kibana
• DaemonSets - manages filebeat(or anything that needs to run on all kubernetes hosts)
• Secrets - list of secrets like passwords,certificates,etc
• Services - list of services
• ...

• If you want to run individual jobs all eck_*_1.sh, scripts can be ran with eck_*_1.sh /path/file.json Example: /path/eck_beat_1.sh beat.json

• If you want to run individual jobs for all eck_*_2.sh, scripts can be ran with eck_*_2.sh /path/proper.json resourcename Example: /path/eck_beat_2.sh beat.json filebeat

After the eck-glance runs I would start looking at eck_events.txt from the <namespace> directory, then eck_pods.txt, etc.. depending on the issue at hand.

• Find # FIX for things to fix.
• Validate code against different diags to ensure proper coverage
• Find ways to remove loops
• Find ways to make current jq queries more simple and account for various nulls