Revised how wait-for-running was done in AWS deployment

deploy-cluster-aws/get_instance_status.py

@@ -1,99 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-
-import json
-import os
-import boto.ec2
-import time
-import argparse
-
-
-AWS_ACCESS_KEY_ID = os.environ["AWS_ACCESS_KEY_ID"]
-AWS_SECRET_ACCESS_KEY = os.environ["AWS_SECRET_ACCESS_KEY"]
-
-parser = argparse.ArgumentParser()
-parser.add_argument("--tag", help="tag instances in aws")
-args = parser.parse_args()
-
-if args.tag:
-    tag = args.tag
-else:
-    # reading credentials from config for remote connection
-    print('usage: python3 get_instance_status.py --tag <tag>')
-    print('reason: tag missing!!!')
-    exit(1)
-
-conn = boto.ec2.connect_to_region("eu-central-1",
-            aws_access_key_id=AWS_ACCESS_KEY_ID,
-            aws_secret_access_key=AWS_SECRET_ACCESS_KEY)
-
-PASSING = []
-RUNNING = []
-
-
-# get list of instance ids from amazon
-def list_of_ids(tag):
-    # TODO: CHANGE TO PROPER DOCSTRING
-    # Returns a list of ids of all instances with the given tag
-    reservations = conn.get_all_instances(filters={"tag:Name": tag})
-    # There are several reservations
-    # and each reservation can have several instances
-    id_list = []
-    for reservation in reservations:
-        for instance in reservation.instances:
-            if instance.id is not None:
-                id_list.append(instance.id)
-    return id_list
-
-
-# Andreas' old code:
-"""
-INSTANCE_IDS = []
-
-def prepare_list(tag):
-    reservations = conn.get_all_instances(filters={"tag:Name" : tag})
-    instances = [i for r in reservations for i in r.instances]
-    for i in instances:
-        inst = i.__dict__
-        inst_id = inst.get('id')
-        INSTANCE_IDS.append(inst_id)
-    return INSTANCE_IDS
-"""
-
-
-# get statuses from amazon
-def create_run_pass_list(tag):
-    # instlist_old = prepare_list(tag)
-    # print("instlist_old = {}".format(instlist_old))
-    instlist_new = list_of_ids(tag)
-    print("instlist_new = {}".format(instlist_new))
-
-    instlist = instlist_new
-
-    for entry in range(0, len(instlist)):  # 0, 1, ..., [len(instlist) - 1]
-        instances = conn.get_only_instances(instance_ids=instlist[entry])
-        status = conn.get_all_instance_status(instance_ids=instlist[entry])
-        for instance in instances:
-            reachability = status[0].system_status.details["reachability"]
-            PASSING.append(reachability)
-    return instlist, PASSING, RUNNING
-
-
-if __name__ == "__main__":
-    # get lists from amazon
-    try:
-        instlist, passlist, runlist = create_run_pass_list(tag)
-        print("instlist = {}".format(instlist))
-        print("passlist = {}".format(passlist))
-        print("runlist = {}".format(runlist))
-    except IndexError:
-        print("Searching for matching cluster-tag...")
-        exit(1)
-
-    for entry in range(0,len(instlist)):
-        if "passed" in passlist and len(set(passlist)) == 1:
-            print("up and running")
-            exit(0)
-
-    # exit with error code for continous check if nothing found
-    exit(1)

deploy-cluster-aws/startup.sh

@@ -1,5 +1,8 @@
 #! /bin/bash
 
+# The set -e option instructs bash to immediately exit if any command has a non-zero exit status
+set -e
+
 function printErr()
     {
         echo "usage: ./startup.sh <tag> <number_of_nodes_in_cluster>"
@@ -51,18 +54,13 @@ fi
 # 0400 for pem/bigchaindb.pem
 chmod 0400 pem/bigchaindb.pem
 
-# starting and tagging instances
+echo "Starting and tagging instances on EC2"
 python run_and_tag.py --tag $TAG --nodes $NODES
 # let's wait a minute to get the nodes ready and in status initializing
 #sleep 60
 
-# checking if instances are up and running (every 5 secs.)
-RET=1
-until [ ${RET} -eq 0 ]; do
-    python3 get_instance_status.py --tag $TAG
-    RET=$?
-    sleep 5
-done
+# Wait until all those instances are running
+python wait_until_all_running.py --tag $TAG
 
 # in case of elastic ips...
 python3 get_elastic_ips.py --tag $TAG

deploy-cluster-aws/wait_until_all_running.py

@@ -0,0 +1,45 @@
+# -*- coding: utf-8 -*-
+
+import os
+import boto3
+import argparse
+
+
+AWS_ACCESS_KEY_ID = os.environ['AWS_ACCESS_KEY_ID']
+AWS_SECRET_ACCESS_KEY = os.environ['AWS_SECRET_ACCESS_KEY']
+AWS_REGION = os.environ['AWS_REGION']
+
+parser = argparse.ArgumentParser()
+parser.add_argument("--tag", help="tag instances in aws")
+args = parser.parse_args()
+
+if args.tag:
+    tag = args.tag
+else:
+    # reading credentials from config for remote connection
+    print('usage: python get_instance_status.py --tag <tag>')
+    print('reason: tag missing!!!')
+    exit(1)
+
+# Connect to Amazon EC2
+ec2 = boto3.resource(service_name='ec2',
+                     region_name=AWS_REGION,
+                     aws_access_key_id=AWS_ACCESS_KEY_ID,
+                     aws_secret_access_key=AWS_SECRET_ACCESS_KEY)
+
+# Get a list of all instances with the specified tag.
+# (Technically, instances_with_tag is an ec2.instancesCollection.)
+instances_with_tag = ec2.instances.filter(
+    Filters=[{'Name': 'tag:Name', 'Values': [tag]}]
+    )
+print('The instances with tag {} have these ids:'.format(tag))
+for instance in instances_with_tag:
+    print(instance.id)
+
+print('Waiting until all those instances exist...')
+for instance in instances_with_tag:
+    instance.wait_until_exists()
+
+print('Waiting until all those instances are running...')
+for instance in instances_with_tag:
+    instance.wait_until_running()