Key Features • Installation • How To Use • Examples • License • Developing
Esctl is a CLI tool for Elasticsearch. I designed it to shorten huge curl
commands Elasticsearch operators were running like :
curl -XPUT --user "john:doe" 'http://elasticsearch.example.com:9200/_cluster/settings' -d '{
"transient" : {
"cluster.routing.allocation.enable": "NONE"
}
}'
The equivalent with esctl
is
esctl cluster routing allocation enable none
- Easy to use CLI rather than long curl commands (thanks to cliff)
- Cluster-level informations : stats, info, health, allocation explanation
- Node-level informations : list, hot threads, exclusion, stats
- Cluster-level and index-level settings
_cat
API for allocation, plugins and thread pools- Index management : open, close, create, delete, list
raw
command to perform raw HTTP calls when esctl doesn't provide a nice interface for a given route.- Per-module log configuration
- X-Pack APIs : users and roles
- Multiple output formats : table, csv, json, value, yaml
- JMESPath queries using the
--jmespath
flag - Colored output !
- Run arbitrary pre-commands before issuing the call to Elasticsearch (like running
kubectl port-forward
for example) - Fetch cluster's credentials from external commands instead of having them shown in cleartext in the config file
pip install esctl
pip install git+https://github.com/jeromepin/esctl.git
Esctl relies on a ~/.esctlrc
file containing its config. This file is automatically created on the first start if it doesn't exists :
clusters:
bar:
servers:
- https://bar.example.com
users:
john-doe:
username: john
external_password:
command:
run: kubectl --context=bar --namespace=baz get secrets -o json my-secret | jq -r '.data.password||@base64d'
contexts:
foo:
user: john-doe
cluster: bar
default-context: foo
Sometimes, you need to execute a shell command right before running the esctl
command. Like running a kubectl port-forward
in order to connect to your Kubernetes cluster.
There is a pre_commands
block inside the context which can take care of that :
clusters:
remote-kubernetes:
servers:
- http://localhost:9200
contexts:
my-distant-cluster:
cluster: remote-kubernetes
pre_commands:
- command: kubectl --context=my-kubernetes-context --namespace=elasticsearch port-forward svc/elasticsearch 9200
wait_for_exit: false
wait_for_output: Forwarding from
user: john-doe
Along with command
, you can pass two options :
wait_for_exit
(default:true
) : wait for the command to exit before continuing. Usually set tofalse
when the command is running in the foreground.wait_for_output
: ifwait_for_exit
isfalse
, look for a specific output in the command's stdout. The string to look-for is interpreted as a regular expression passed to Python's re.compile().
esctl
is licensed under the GNU GPLv3. See LICENCE file.
make install
make test
make lint