I'd like to announce es-diag
I've decided to build a tool for our ops, to assist discovering improper
Elastic Search machine configuration.
The goal is to provide 0-friction and fast problem discovery for ops before
they start implementing Elastic Search,
and while they are running an instance.
Right now the bare minimum is implemented in terms of checks - that is, the
common 'too many open files' ulimit problem.
However the needed framework is in place such that you can write a 'check'
script using the simple DSL-like structure, and send a pull request pretty
In the immediate future, I'm planning integration with a running instance's
cluster info to warn about health problems.
In addition, a quiet-mode where the tool can be easily integrated with
monitoring tools such as nagios/zabbix and the sorts.
Please feel free to ask me questions directly, here, or open issues/pull
Hope this helps,