ecstats is the collective name of scripts (pullElasticCacheStats.py, pullAzureCacheForRedisStats.py, pullRedisOpenSourceStats.py) which will pull raw usage data from ElastiCache, Azure Cache for Redis and a Redis Open Source cluster, respectively.

The output from these scripts are then analyzed using the tools in the (private to Redislabs) EC2RL-Internal toolset.

We assume you have the appropriate CLI installed (AWS, Azure) for your cloud provider and have appropriate authorization and authentication, or know the host/port/password if connecting to an OSS cluster.

The easiest way forward is if you have Docker installed.

For Azure, having jq installed is helpful too (although if you've got Docker we'll assume you can use the jq image for docker)

The simplest mechanism is to use the ecstats Docker image.

Each script is a little different:

Executing this command is the default command in the docker file.

$ cp config.cfg.example config.cfg

$ docker run -v $PWD:/ecstats docker.pkg.github.com/redislabs-solution-architects/ecstats/ecstats:latest

Note that we mount the current directory onto /ecstats. This name ecstats is built into the default CMD, so go with it!

Results will be stored in the mounted folder. For each section in the config.cfgfile there'll be a an Excel file generated in the/ecstats` directory (you did remember to mount that onto your current directory, didn't you?!)

# Clone:
git clone https://github.com/Redislabs-Solution-Architects/ecstats

# Prepare virtualenv:
cd ecstats
mkdir .env
virtualenv .env

# Activate virtualenv
. .env/bin/activate

# Install necessary libraries
pip install -r requirements.txt

# When finished
deactivate

To run the script copy the configuration file and edit:

cp config.cfg.example config.cfg

Execute

python pullElasticCacheStats.py -c config.cfg

The output will be in a file called AzureStats.xlsx in the current directory.

You need to authenticate to Azure and setup an MSI first, before running the actual script. I've automated as much of this as possible, but if you don't want to run jq then simply run the az ad ... part of the command and create the necessary envars by hand.

az login
$(az ad sp create-for-rbac --name http://pullAzureStats  | docker run -i stedolan/jq -r '"export AZURE_CLIENT_ID=" + .appId + " AZURE_CLIENT_SECRET=" + .password + " AZURE_TENANT_ID=" +.tenant')
docker run -e AZURE_TENANT_ID -e AZURE_CLIENT_ID -e AZURE_CLIENT_SECRET -v$(pwd):/ecstats docker.pkg.github.com/redislabs-solution-architects/ecstats/ecstats:latest python /app/pullAzureCacheForRedisStats.py -d /ecstats

The second line above sets up all the necessary envars.

The -d /ecstats argument in the third line sets up the output directory to the directory mounted onto the current directory (so the output file will be available afterwards!)

Follow the instructions for the virtual environment above, then execute the following:

az login
$(az ad sp create-for-rbac --name http://pullAzureStats  | jq -r '"export AZURE_CLIENT_ID=" + .appId + " AZURE_CLIENT_SECRET=" + .password + " AZURE_TENANT_ID=" +.tenant')
python pullAzureCacheForRedisStats.py

The script extracts the current cluster usage by using the INFO and INFO COMMANDSTATS commands. These two Redis commands are called twice, in order to measure in order to capture the commands process during this timeframe.

The script takes as an input the following:

• an Excel files with the Redis DB access configuration The Excel contains the following columns:

DB Name - a logical name for the Redis DB Redis Host - the host/IP address of the DB Port - DB port Password - the DB password User (ACL) - a username to connect with to the DB which has privileges to the INFO command

A template can be found in: samples/sampleOSSPullInput.xlsx.

$ docker run -v $PWD:/ecstats docker.pkg.github.com/redislabs-solution-architects/ec2rl-internal/ec2rl-internal:latest python pullRedisOpenSourceStats.py /ecstats/sampleOSSPullInput.xlsx

python pullRedisOpenSourceStats.py sampleOSSPullInput.xlsx

The output will be an Excel file with all the information gathered from the clusters. An example can be found in samples/sampleOSSStats.xlsx.