tag-old-ebs.py 7.5 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
"""

For a given aws account, go through all un-attached volumes and tag them.

"""
import boto
import boto.utils
import argparse
import logging
import subprocess
import time
import os
from os.path import join, exists, isdir, islink, realpath, basename, dirname
import yaml
# needs to be pip installed
import netaddr

LOG_FORMAT = "%(asctime)s %(levelname)s - %(filename)s:%(lineno)s - %(message)s"
TIMEOUT = 300

log_level = logging.INFO

def tags_for_hostname(hostname, mapping):
    logging.debug("Hostname is {}".format(hostname))
    if not hostname.startswith('ip-'):
        return {}

    octets = hostname.lstrip('ip-').split('-')
    tags = {}

    # Update with env and deployment info
    tags.update(mapping['CIDR_SECOND_OCTET'][octets[1]])

    ip_addr = netaddr.IPAddress(".".join(octets))
    for key, value in mapping['CIDR_REST'].items():
        cidr = ".".join([
            mapping['CIDR_FIRST_OCTET'],
            octets[1],
            key])

        cidrset = netaddr.IPSet([cidr])

        if ip_addr in cidrset:
            tags.update(value)

    return tags

def potential_devices(root_device):
    device_dir = dirname(root_device)
    relevant_devices = lambda x: x.startswith(basename(root_device))

    all_devices = os.listdir(device_dir)
    all_devices = filter(relevant_devices, all_devices)

    logging.info("Potential devices on {}: {}".format(root_device, all_devices))
    if len(all_devices) > 1:
        all_devices.remove(basename(root_device))

    return map(lambda x: join(device_dir, x), all_devices)

def get_tags_for_disk(mountpoint):
    tag_data = {}
    # Look at some files on it to determine:
    #  - hostname
    #  - environment
    #  - deployment
    #  - cluster
    #  - instance-id
    #  - date created
    hostname_file = join(mountpoint, "etc", "hostname")
    edx_dir = join(mountpoint, 'edx', 'app')
    if exists(hostname_file):
        # This means this was a root volume.
        with open(hostname_file, 'r') as f:
            hostname = f.readline().strip()
            tag_data['hostname'] = hostname

        if exists(edx_dir) and isdir(edx_dir):
            # This is an ansible related ami, we'll try to map
            # the hostname to a knows deployment and cluster.
            cluster_tags = tags_for_hostname(hostname, mappings)
            tag_data.update(cluster_tags)
        else:
            # Not an ansible created root volume.
            tag_data['cluster'] = 'unknown'
    else:
        # Not a root volume
        tag_data['cluster'] = "unknown"

    instance_file = join(mountpoint, "var", "lib", "cloud", "instance")
    if exists(instance_file) and islink(instance_file):
        resolved_path = realpath(instance_file)
        old_instance_id = basename(resolved_path)
        tag_data['instance-id'] = old_instance_id

    return tag_data


if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Tag unattached ebs volumes.")
    parser.add_argument("--profile", '-p',
        help="AWS Profile to use with boto.")
    parser.add_argument("--noop", "-n", action="store_true",
        help="Don't actually tag anything.")
    parser.add_argument("--verbose", "-v", action="store_true",
        help="More verbose output.")
    parser.add_argument("--device", "-d", default="/dev/xvdf",
        help="The /dev/??? where the volume should be mounted.")
    parser.add_argument("--mountpoint", "-m", default="/mnt",
        help="Location to mount the new device.")
    parser.add_argument("--config", "-c", required=True,
        help="Configuration to map hostnames to tags.")
    # The config should specify what tags to associate with the second
    # and this octet of the hostname which should be the ip address.
    # example:

    args = parser.parse_args()

    mappings = yaml.safe_load(open(args.config,'r'))

    # Setup Logging
    if args.verbose:
        log_level = logging.DEBUG

    logging.basicConfig(format=LOG_FORMAT, level=log_level)

    # setup boto
    ec2 = boto.connect_ec2(profile_name=args.profile)

    # get mounting args
    id_info = boto.utils.get_instance_identity()['document']
    instance_id = id_info['instanceId']
    az = id_info['availabilityZone']
    root_device = args.device
    mountpoint = args.mountpoint

    # Find all unattached volumes
    filters = { "status": "available", "availability-zone": az }
    potential_volumes = ec2.get_all_volumes(filters=filters)
    logging.debug("Found {} unattached volumes in {}".format(len(potential_volumes), az))

    for vol in potential_volumes:
        if "cluster" in vol.tags:
            continue
        # Attach volume to the instance running this process
        logging.debug("Trying to attach {} to {} at {}".format(
            vol.id, instance_id, root_device))

        try:
            ec2.attach_volume(vol.id, instance_id, root_device)
            # Wait for the volume to finish attaching.
            waiting_msg = "Waiting for {} to be available at {}"
            timeout = TIMEOUT
            while not exists(root_device):
                time.sleep(2)
                logging.debug(waiting_msg.format(vol.id, root_device))
                timeout -= 2
                if timeout <= 0:
                    logging.critical("Timed out while attaching {}.".format(vol.id))
                    exit(1)


            # Because a volume might have multiple mount points
            devices_on_volume = potential_devices(root_device)
            if len(devices_on_volume) != 1:
                vol.add_tag("devices_on_volume", str(devices_on_volume))
                # Don't tag in this case because the different devices
                # may have conflicting tags.
                logging.info("Skipping {} because it has multiple mountpoints.".format(vol.id))
                logging.info("{} has mountpoints {}".format(vol.id, str(devices_on_volume)))
            else:
                device = devices_on_volume[0]
                try:
                    # Mount the volume
                    subprocess.check_call(["sudo", "mount", device, mountpoint])

                    # Learn all tags we can know from content on disk.
                    tag_data = get_tags_for_disk(mountpoint)
                    tag_data['created'] = vol.create_time

                    # If they are found tag the instance with them
                    if args.noop:
                        logging.info("Would have tagged {} with: \n{}".format(vol.id, str(tag_data)))
                    else:
                        logging.info("Tagging {} with: \n{}".format(vol.id, str(tag_data)))
                        vol.add_tags(tag_data)
                finally:
                    # Un-mount the volume
                    subprocess.check_call(['sudo', 'umount', mountpoint])
        finally:
            # Need this to be a function so we always re-check the API for status.
            is_attached = lambda vol_id: ec2.get_all_volumes(vol_id)[0].status != "available"
            timeout = TIMEOUT
            while exists(root_device) or is_attached(vol.id):
                if is_attached(vol.id):
                    try:
                        # detach the volume
                        ec2.detach_volume(vol.id)
                    except boto.exception.EC2ResponseError as e:
                        logging.warning("Failed to detach volume.  Will try again in a bit.")
                time.sleep(2)
                timeout -= 2
                if timeout <= 0:
                    logging.critical("Timed out while detaching {}.".format(vol.id))
                    exit(1)
                logging.debug("Waiting for {} to be detached.".format(vol.id))