chrisdpa-tvx

set -exu
SPOTFLEETREQ=$1
CLUSTER=$2
AZ=$3
fleet_sz=$(aws ec2 describe-spot-fleet-instances --spot-fleet-request-id ${SPOTFLEETREQ} | jq -r '.ActiveInstances|length')

container_sz=$(aws ecs list-container-instances --cluster ${CLUSTER} | jq -r '.containerInstanceArns[]' | \
xargs -n10 -J % aws ecs describe-container-instances --cluster ${CLUSTER} --container-instances % | \
jq -r ".containerInstances[].attributes[]|select(.name==\"ecs.availability-zone\" and .value==\"${AZ}\")|.value"|wc -l | xargs) 

chrisdpa-tvx

CLUSTER=$1

aws ecs list-services --cluster ${CLUSTER} | jq -r '.[] | .[]' | \
xargs -J % -n10 aws ecs describe-services --cluster ${CLUSTER} --services % | jq -r '.services[] | .serviceArn' | \
xargs -J % -n10 aws ecs describe-services --cluster ${CLUSTER} --services % | jq -r '.services[]|.taskDefinition' | \
xargs -J % -n1 aws ecs describe-task-definition --task-definition % | jq -r '.taskDefinition|.containerDefinitions[]|.image'

chrisdpa-tvx

CLUSTER=$1
sts aws ecs list-services --cluster ${CLUSTER} |\
jq -r '.[] | .[]' | xargs -J % -n10 sts \
aws ecs describe-services --cluster ${CLUSTER} \
--services % | jq -r '.services[] | select(.desiredCount!=.runningCount) | .serviceArn' > services

while read r;do 
  sts aws ecs update-service \
  --cluster ${CLUSTER} \
  --service ${r} \

chrisdpa-tvx

See netstat detail of a running docker container:

nsenter -t $(docker inspect -f '{{.State.Pid}}' _container_id_) -n netstat -tunapp

chrisdpa-tvx

All Your Data Does Not Belong In a Database

Businesses are machines producing mountains of data about sales, usage, customer, costs, etc... Traditionally data processing is highly centralised with teams of staff and computer running hot a whirling ready to process. We can do better than moving the mountain of data into the corporate data machine - so long as that machinary is light enough to be moved to the data.

Don't move the mountain - Bring the processing to the data

We've had this problem; a huge directory of files in CSV format, conataining vital information for our business. But it's in CSV, requires analysis, and don't you don't feel like learning sed/grep/awk today - besides it's 2017 and no-one thinks those tools are easy to use.

chrisdpa-tvx

import gnupg
import cStringIO

gpg = gnupg.GPG()
gpg.encoding = 'utf-8'

me = {'name_real': 'alice',
      'name_email': 'alice@example.com',
      'expire_date': '2024-04-01',
      'passphrase': 'alicespassword'}

chrisdpa-tvx

If you don't have dos2unix installed use sed

dos2unix: sed 's/\r//'

unix2dos: sed 's/$/\r/'

https://github.com/sindresorhus/awesome