Skip to content

Latest commit

 

History

History
830 lines (625 loc) · 38.2 KB

README.md

File metadata and controls

830 lines (625 loc) · 38.2 KB

aws-management-suite

Source URL: https://github.com/dwayn/aws-management-suite

Introduction

This is currently very much a work in progress, and there is much that will be cleaned up over time. The goal of this suite is to abstract many of the common tasks related to managing cloud infrastructure in AWS. The primary goal of this tool is to bridge the gap between raw infrastructure management tools like the EC2 command line tools, and configuration management tools. Initially, this tool is focused on EBS volume, raid and snapshot management, but going forward the goal is to cover other infrastructure management needs that are not fully addressed by other tools.

Current Features

SSH client

  • password or private key based login
  • support for sudo login (password or passwordless)
  • captures stdout, stderr and exit code from command run

EBS Volumes (managed as groups of volumes)

  • create volume/raid
  • delete volume/raid
  • attach volume/raid
  • detach volume/raid
  • create software raid volume
  • partition and format new volumes/raids
  • (re)assemble software raid from existing (or newly cloned) ebs volumes
  • mount volume/raid
  • unmount volume/raid

EBS Snapshots (managed as groups of snapshots)

  • pre/post snapshot hooks to enable running commands/scripts on target host before and after starting snapshot to ensure consistent point in time snapshot of all volumes in a raid group
  • copy snapshot group to another region (only handled internally currently)
  • clone snapshot group to new volume group and optionally attach/mount on a host
  • clone latest snapshot of a volume group or host/instance + mount point
  • schedule regular snapshots of volume/raid with managed grandfather/father/son expiration
  • automatable purging of expired snapshots

Instance Management

  • Instances can be manually added or edited using ams host add or ams host edit respectively
  • Instance discovery has been implemented, allowing the hosts table to be automatically populated

Setup and Configuration

Initial installation

  • Copy sample_settings.py to settings.py and edit AWS, SSH and SUDO access credentials
  • A MySQL database needs to be setup for tracking state. The following statements assume that the mysql database and the tool are located on the same host:
  • CREATE DATABASE ams; -- Create the schema
  • GRANT ALL PRIVILEGES ON ams.* TO 'ams_user'@'localhost' IDENTIFIED BY 'mypass'; -- (This will create user with username 'ams_user' and password of 'mypass' and give access to the new schema created)
  • Edit TRACKING_DB credentials in settings.py with the proper credentials for your MySQL database
  • pip install -r requirements.txt will install the handful of external dependencies
  • Suggested: add the path to ams directory to your path or add symlink to ams script to a directory in the system path
  • ams internals database install will create the current full version of all of the tables

Enabling bash/zsh completion

This project makes use of the argcomplete library (https://github.com/kislyuk/argcomplete) to provide dynamic completion. As part of the pip installation, the library will be installed, but completion will still need to be enabled. Due to some multi-platform issues I experienced trying to enable global completeion, I opted to use specific completion. All that is needed is to add this line to your .bashrc, .profile or .bash_profile (depending on which your OS uses) and then reload your terminal or source .bashrc (or .profile or .bash_profile).

eval "$(register-python-argcomplete ams)"

Upgrading

If you have updated the code base, just run pip install -r requirements.txt to install any new dependencies and run ams internals database upgrade to run the update scripts for the database tables. Upgrade scripts can be run as often as you like, as it will do nothing if the database version matches the code version.

Management Tool Usage

General

All of the functionality is through the command line tool ams. It has been implemented as a multi-level nested command parser using the argparse module.
If at any point you need help just add -h or --help flag to the command line and it will list all available sub-commands and options for the current command level.
There are still a few legacy command structures that need to be cleaned up, so there may be some minor changes to the syntax to a few of these, but I will attempt to keep these to an absolute minimum.

Host/Instance

ams host list

With no options this lists all host entries in the database

Arguments:

  --zone ZONE         Availability zone to filter results by. This is a prefix
                      search so any of the following is valid with increasing
                      specificity: 'us', 'us-west', 'us-west-2', 'us-west-2a'

ams host list host [hostname]

If hostname is given then it will match hostname exactly

Arguments:

  --like LIKE         wildcard matches hostname
  --prefix PREFIX     prefix matches hostname
  --zone ZONE         Availability zone to filter results by. This is a prefix
                      search so any of the following is valid with increasing
                      specificity: 'us', 'us-west', 'us-west-2', 'us-west-2a'

ams host list instance_id [instance id]

If instance id is given then it will match instance_id exactly

Arguments:

  --like LIKE         wildcard matches instance id
  --prefix PREFIX     prefix matches instance id
  --zone ZONE         Availability zone to filter results by. This is a prefix
                      search so any of the following is valid with increasing
                      specificity: 'us', 'us-west', 'us-west-2', 'us-west-2a'

ams host add

Add a host to the hosts table so that resources on the host can be managed. This has effectively been replaced by the host discovery functionality.

Required arguments: --instance, --host, --zone

Arguments:

  -i INSTANCE, --instance INSTANCE
                        Instance ID of the instance to add
  -H HOSTNAME, --hostname HOSTNAME
                        hostname of the host (used to ssh to the host to do
                        management)
  -z ZONE, --zone ZONE  availability zone that the instance is in
  --hostname-internal HOSTNAME_INTERNAL
                        internal hostname (stored but not currently used)
  --hostname-external HOSTNAME_EXTERNAL
                        external hostname (stored but not currently used)
  --ip-internal IP_INTERNAL
                        internal IP address (stored but not currently used)
  --ip-external IP_EXTERNAL
                        external IP address (stored but not currently used)
  --ami-id AMI_ID       AMI ID (stored but not currently used)
  --instance-type INSTANCE_TYPE
                        Instance type (stored but not currently used)
  --notes NOTES         Notes on the instance/host (stored but not currently
                        used)
  -z ZONE, --zone ZONE  availability zone that the instance is in

ams host edit

Edit a host's details in the database, particularly useful for editing the hostname which does not get overwritten on discovery passes.

Required arguments: --instance

Arguments:

  -i INSTANCE, --instance INSTANCE
                        Instance ID of the instance to add
  -H HOSTNAME, --hostname HOSTNAME
                        hostname of the host (used to ssh to the host to do
                        management)
  -z ZONE, --zone ZONE  availability zone that the instance is in
  --hostname-internal HOSTNAME_INTERNAL
                        internal hostname (stored but not currently used)
  --hostname-external HOSTNAME_EXTERNAL
                        external hostname (stored but not currently used)
  --ip-internal IP_INTERNAL
                        internal IP address (stored but not currently used)
  --ip-external IP_EXTERNAL
                        external IP address (stored but not currently used)
  --ami-id AMI_ID       AMI ID (stored but not currently used)
  --instance-type INSTANCE_TYPE
                        Instance type (stored but not currently used)
  --notes NOTES         Notes on the instance/host (stored but not currently
                        used)
  -z ZONE, --zone ZONE  availability zone that the instance is in

ams host discovery

Runs host discovery to populate the hosts table automatically

Arguments: None

Volumes

ams volume list

With no options this lists all volume groups in the database

Arguments:

  --zone ZONE         Availability zone to filter results by. This is a prefix
                      search so any of the following is valid with increasing
                      specificity: 'us', 'us-west', 'us-west-2', 'us-west-2a'

ams volume list host [hostname]

Lists the volume groups for a host or hosts
If hostname is given then it will match hostname exactly

Arguments:

  --like LIKE         wildcard matches hostname
  --prefix PREFIX     prefix matches hostname
  --zone ZONE         Availability zone to filter results by. This is a prefix
                      search so any of the following is valid with increasing
                      specificity: 'us', 'us-west', 'us-west-2', 'us-west-2a'

ams volume list instance_id [instance_id]

Lists the volume groups for an instance or instances
If instance id is given then it will match instance_id exactly

Arguments:

  --like LIKE         wildcard matches instance id
  --prefix PREFIX     prefix matches instance id
  --zone ZONE         Availability zone to filter results by. This is a prefix
                      search so any of the following is valid with increasing
                      specificity: 'us', 'us-west', 'us-west-2', 'us-west-2a'

ams volume create

Creates a new volume group (single or multiple disk) and attaches to host. Optionally mounts the volume and configures automounting.

Required arguments: (--host | --instance), --numvols, --size

Defaults:

  • stripe-block-size: 256 (256k chunk size recommended for performance of EBS stripes using xfs)
  • raid-level: 0
  • filesystem: xfs (note: currently due to implementation constrictions filesystem must be one of the types that can be formatted using mkfs.*)
  • iops: None
  • mount-point: None (disk will not be mounted and automounting will not be configured if mount-point not provided)
  • no-automount: false (automounting of volumes/raids will be configured in fstab and mdadm.conf by default unless explicitly disabled)

Arguments:

  -i INSTANCE, --instance INSTANCE
                        instance_id of an instance to attach new volume group
  -H HOST, --host HOST  hostname of an instance to attach new volume group
  -n NUMVOLS, --numvols NUMVOLS
                        Number of EBS volumes to create for the new volume
                        group
  -r {0,1,5,10}, --raid-level {0,1,5,10}
                        Set the raid level for new EBS raid
  -b STRIPE_BLOCK_SIZE, --stripe-block-size STRIPE_BLOCK_SIZE
                        Set the stripe block/chunk size for new EBS raid
  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        Set the mount point for volume. Not required, but
                        suggested
  -a, --no-automount    Disable configuring the OS to automatically mount the
                        volume group on reboot
  -f FILESYSTEM, --filesystem FILESYSTEM
                        Filesystem to partition new raid/volume
  -s SIZE, --size SIZE  Per EBS volume size in GiBs
  -p IOPS, --iops IOPS  Per EBS volume provisioned iops

ams volume delete (volume_group_id)

Deletes provided volume_group_id. Volume group must not be currently attached to an instance.

Required arguments: volume_group_id

ams volume attach (volume_group_id)

Attaches provided volume_group_id to a host. Optionally mounts the volume and configures automounting.

Required arguments: volume_group_id, (--host | --instance)

Defaults:

  • mount-point: None (disk will not be mounted and automounting will not be configured if mount-point not provided)
  • no-automount: false (automounting of volumes/raids will be configured in fstab and mdadm.conf by default unless explicitly disabled)

Arguments:

  -i INSTANCE, --instance INSTANCE
                        instance_id of an instance to attach new volume group
  -H HOST, --host HOST  hostname of an instance to attach new volume group
  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        Set the mount point for volume. Not required, but
                        suggested
  -a, --no-automount    Disable configuring the OS to automatically mount the
                        volume group on reboot

ams volume detach volume (volume_group_id)

Detaches provided volume_group_id from the host it is currently attached. Removes the automounting configuration for the volume group.

Required arguments: volume_group_id

Arguments:

  -u, --unmount         Unmounts the volume group if it is mounted. If this
                        option is not included and the volume is mounted the
                        detach operation will fail
  -f FORCE, --force FORCE
                        Force detach the volume group's EBS volumes

ams volume detach host (hostname) (mount_point)

Detaches provided volume group that is mounted at mount_point on hostname. Removes the automounting configuration for the volume group.

Required arguments: hostname, mount_point

Arguments:

  -u, --unmount         Unmounts the volume group if it is mounted. If this
                        option is not included and the volume is mounted the
                        detach operation will fail
  -f FORCE, --force FORCE
                        Force detach the volume group's EBS volumes

ams volume detach instance (instance_id) (mount_point)

Detaches provided volume group that is mounted at mount_point on instance_id. Removes the automounting configuration for the volume group.

Required arguments: instance_id, mount_point

Arguments:

  -u, --unmount         Unmounts the volume group if it is mounted. If this
                        option is not included and the volume is mounted the
                        detach operation will fail
  -f FORCE, --force FORCE
                        Force detach the volume group's EBS volumes

ams volume mount (volume_group_id)

Mount a volume group on the host that it is currently attached. Supports mounting to a given mount point or the currently defined mount point for the volume group.

Required arguments: volume_group_id

Arguments:

  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        Set the mount point for volume. If not provided, will
                        attempt to use currently defined mount point
  -a, --no-automount    Disable configure the OS to automatically mount the
                        volume group on reboot

ams volume unmount (volume_group_id)

Unmount volume_group_id on the host that it is currently mounted. Does not make any changes to currently automount configuration.

Required arguments: volume_group_id

ams volume automount (volume_group_id)

Configure automounting for the volume_group_id. If mount point is not provided then it will use the currently defined mount point for the volume. If neither of these exist then it will configure automounting of the volume where it is currently mounted, otherwise it will fail configuring automounting.

Required arguments: volume_group_id

Arguments:

  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        Set the mount point for volume. If not provided, will
                        attempt to use currently defined mount point
  -r, --remove          Remove the current automount configuration for a
                        volume group

Snapshots

ams snapshot list volume (volume_group_id)

List the snapshots of a specific volume_group_id.

Required arguments: volume_group_id

Arguments:

  -r REGION, --region REGION
                        Filter the snapshots by region

ams snapshot list host [hostname]

List the snapshots for a specific host, or for hosts matching a search string. Optionally filter by mount point and/or region.

Arguments:

  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        Filter the snapshots by the mount point
  -r REGION, --region REGION
                        Filter the snapshots by region
  --like LIKE           search string to use to filter hosts
  --prefix PREFIX       search string prefix to filter hosts

ams snapshot list instance [instance_id]

List the snapshots for a specific instance, or for instances matching a search string. Optionally filter by mount point and/or region.

Arguments:

  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        Filter the snapshots by the mount point
  -r REGION, --region REGION
                        Filter the snapshots by region
  --like LIKE           search string to use to filter hosts
  --prefix PREFIX       search string prefix to filter hosts

ams snapshot create volume (volume_group_id)

Create a snapshot of a specific volume_group_id.
PRE and POST are commands that will be run before and after the snapshot, and provide a means to ensure that data is in a consistent state before snapshotting and revert back to normal operation after snapshot has begun. Description is written as metadata to the snapshot itself and will show up in the EC2 console.

Required arguments: volume_group_id

Arguments:

  --pre PRE             command to run on host to prepare for starting EBS
                        snapshot (will not be run if volume group is not
                        attached)
  --post POST           command to run on host after snapshot (will not be run
                        if volume group is not attached)
  -d DESCRIPTION, --description DESCRIPTION
                        description to add to snapshot(s)

ams snapshot create host

Create a snapshot of a specific volume that is on a host
PRE and POST are commands that will be run before and after the snapshot, and provide a means to ensure that data is in a consistent state before snapshotting and revert back to normal operation after snapshot has begun.
Description is written as metadata to the snapshot itself and will show up in the EC2 console.

Required arguments: (--host | --instance), --mount-point

Arguments:

  -i INSTANCE, --instance INSTANCE
                        instance_id of an instance to snapshot a volume group
  -H HOST, --host HOST  hostname of an instance to snapshot a volume group
  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        mount point of the volume group to snapshot
  --pre PRE             command to run on host to prepare for starting EBS
                        snapshot (will not be run if volume group is not
                        attached)
  --post POST           command to run on host after snapshot (will not be run
                        if volume group is not attached)
  -d DESCRIPTION, --description DESCRIPTION
                        description to add to snapshot(s)

ams snapshot delete expired

Delete all expired snapshots. This operation is intended to be able to be added to crontab for regular purging of expired snapshots

Required arguments: None

Arguments: None

ams snapshot delete snapshot (snapshot_group_id)

Delete a specific snapshot_group_id. Use one of the snapshot list commands to find a snapshot_group_id.

Required arguments: snapshot_group_id

ams snapshot clone snapshot (snapshot_group_id)

Clone a specific snapshot_group_id into a new volume group and optionally attach and mount the new volume.
This will manage copying snapshot to destination region if the destination region is not the same as where the snapshot group is held.

If iops is provided then the volumes in the new volume group will be created with the provided iops, otherwise the iops of the original volume group for the snapshot will be used. To create the volumes in the new volume group with no iops when the original volume group had iops, pass in 0 for iops to explicitly disable.

Required arguments: snapshot_group_id, (--zone | --host | --instance)

Arguments:

  -z ZONE, --zone ZONE  Availability zone to create the new volume group in
  -i INSTANCE, --instance INSTANCE
                        instance id to attach the new volume group to
  -H HOST, --host HOST  hostname to attache the new volume group to
  -m MOUNT_POINT, --mount_point MOUNT_POINT
                        directory to mount the new volume group to
  -a, --no-automount    Disable configuring the OS to automatically mount the
                        volume group on reboot
  -p PIOPS, --iops PIOPS
                        Per EBS volume provisioned iops. Set to 0 to
                        explicitly disable provisioned iops. If not provided
                        then the iops of the original volumes will be used.

ams snapshot clone latest volume (volume_group_id)

Clone the latest snapshot for a volume_group_id and optionally attach and mount the new volume.
This will manage copying snapshot to destination region if the destination region is not the same as where the snapshot group is held.

If iops is provided then the volumes in the new volume group will be created with the provided iops, otherwise the iops of the original volume group for the snapshot will be used. To create the volumes in the new volume group with no iops when the original volume group had iops, pass in 0 for iops to explicitly disable.

Required arguments: volume_group_id, (--zone | --host | --instance)

Arguments:

  -z ZONE, --zone ZONE  Availability zone to create the new volume group in
  -i INSTANCE, --instance INSTANCE
                        instance id to attach the new volume group to
  -H HOST, --host HOST  hostname to attache the new volume group to
  -m MOUNT_POINT, --mount_point MOUNT_POINT
                        directory to mount the new volume group to
  -a, --no-automount    Disable configuring the OS to automatically mount the
                        volume group on reboot
  -p IOPS, --iops IOPS  Per EBS volume provisioned iops. Set to 0 to
                        explicitly disable provisioned iops. If not provided
                        then the iops of the original volumes will be used.

ams snapshot clone latest host (hostname) (src_mount_point)

Clone the latest snapshot for a host + mount-point and optionally attach and mount the new volume.
This will manage copying snapshot to destination region if the destination region is not the same as where the snapshot group is held.

If iops is provided then the volumes in the new volume group will be created with the provided iops, otherwise the iops of the original volume group for the snapshot will be used. To create the volumes in the new volume group with no iops when the original volume group had iops, pass in 0 for iops to explicitly disable.

Required arguments: hostname, src_mount_point, (--zone | --host | --instance)

Arguments:

  -z ZONE, --zone ZONE  Availability zone to create the new volume group in
  -i INSTANCE, --instance INSTANCE
                        instance id to attach the new volume group to
  -H HOST, --host HOST  hostname to attache the new volume group to
  -m MOUNT_POINT, --mount_point MOUNT_POINT
                        directory to mount the new volume group to
  -a, --no-automount    Disable configuring the OS to automatically mount the
                        volume group on reboot
  -p IOPS, --iops IOPS  Per EBS volume provisioned iops. Set to 0 to
                        explicitly disable provisioned iops. If not provided
                        then the iops of the original volumes will be used.

ams snapshot clone latest instance (instance_id) (src_mount_point)

Clone the latest snapshot for an instance + mount-point and optionally attach and mount the new volume.
This will manage copying snapshot to destination region if the destination region is not the same as where the snapshot group is held.

If iops is provided then the volumes in the new volume group will be created with the provided iops, otherwise the iops of the original volume group for the snapshot will be used. To create the volumes in the new volume group with no iops when the original volume group had iops, pass in 0 for iops to explicitly disable.

Required arguments: instance_id, src_mount_point, (--zone | --host | --instance)

Arguments:

  -z ZONE, --zone ZONE  Availability zone to create the new volume group in
  -i INSTANCE, --instance INSTANCE
                        instance id to attach the new volume group to
  -H HOST, --host HOST  hostname to attache the new volume group to
  -m MOUNT_POINT, --mount_point MOUNT_POINT
                        directory to mount the new volume group to
  -a, --no-automount    Disable configuring the OS to automatically mount the
                        volume group on reboot
  -p IOPS, --iops IOPS  Per EBS volume provisioned iops. Set to 0 to
                        explicitly disable provisioned iops. If not provided
                        then the iops of the original volumes will be used.

ams snapshot schedule list [resource] [resource_id]

Lists the snapshot schedules based on a specific resource id. resource is one of the literals host, instance, volume and resource_id is either hostname, instance_id, or volume_group_d respectively.
If no arguments are provided, then it will list all snapshot schedules

Arguments:

  --like LIKE           search string to use when listing resources
  --prefix PREFIX       search string prefix to use when listing resources

ams snapshot schedule add host (hostname)

Schedule snapshots for a host + mount point. This is the most flexible of the snapshot scheduling methods as it will resolve the host and mount point to do snapshots and will not be affected if the instance or the volume group are changed on a host.
Interval settings affect how often a snapshot is performed/retained. eg. 2 for hourly will take an "hourly" snapshot every other hour, 2 for daily will take a "daily" snapshot every other day.
Retain settings affect how many of each type of snapshot to keep. eg. 24 for hours will keep the last 24 "hourly" snapshots (not necessarily the last 24 hours if "hourly" interval is not 1). Setting the retain value for any of the types to 0 disables that one.
If --intervals or --retentions are set, they will override the single int_* and ret_* arguments
PRE and POST are commands that will be run before and after the snapshot, and provide a means to ensure that data is in a consistent state before snapshotting and revert back to normal operation after snapshot has begun.
Description is written as metadata to the snapshot itself and will show up in the EC2 console.

Required arguments: hostname, --mount-point

Defaults:

  • int_hour 1
  • int_day 1
  • int_week 1
  • int_month 1
  • ret_hour 24
  • ret_day 14
  • ret_week 4
  • ret_month 12
  • ret_year 3

Arguments:

  -i HOUR DAY WEEK MONTH, --intervals HOUR DAY WEEK MONTH
                        Set all intervals at once
  -r HOURS DAYS WEEKS MONTHS YEARS, --retentions HOURS DAYS WEEKS MONTHS YEARS
                        Set all retentions at once
  --int_hour HOURS      hourly interval for snapshots
  --int_day DAYS        daily interval for snapshots
  --int_week WEEKS      weekly interval for snapshots
  --int_month MONTHS    monthly interval for snapshots
  --ret_hour HOURS      number of hourly snapshots to keep
  --ret_day DAYS        number of daily snapshots to keep
  --ret_week WEEKS      number of weekly snapshots to keep
  --ret_month MONTHS    number of monthly snapshots to keep
  --ret_year YEARS      number of yearly snapshots to keep
  --pre PRE_COMMAND     command to run on host to prepare for starting EBS
                        snapshot (will not be run if volume group is not
                        attached)
  --post POST_COMMAND   command to run on host after snapshot (will not be run
                        if volume group is not attached)
  -d DESCRIPTION, --description DESCRIPTION
                        description to add to snapshot
  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        mount point of the volume group to snapshot

ams snapshot schedule add instance (instance_id)

Schedule snapshots for an instance + mount point. This is more flexible than the volume group id based snapshot, but if the instance for a host is replaced, then the snapshot may not be able to run.
Interval settings affect how often a snapshot is performed/retained. eg. 2 for hourly will take an "hourly" snapshot every other hour, 2 for daily will take a "daily" snapshot every other day.
Retain settings affect how many of each type of snapshot to keep. eg. 24 for hours will keep the last 24 "hourly" snapshots (not necessarily the last 24 hours if "hourly" interval is not 1). Setting the retain value for any of the types to 0 disables that one.
If --intervals or --retentions are set, they will override the single int_* and ret_* arguments
PRE and POST are commands that will be run before and after the snapshot, and provide a means to ensure that data is in a consistent state before snapshotting and revert back to normal operation after snapshot has begun.
Description is written as metadata to the snapshot itself and will show up in the EC2 console.

Required arguments: instance_id, --mount-point

Defaults:

  • int_hour 1
  • int_day 1
  • int_week 1
  • int_month 1
  • ret_hour 24
  • ret_day 14
  • ret_week 4
  • ret_month 12
  • ret_year 3

Arguments:

  -i HOUR DAY WEEK MONTH, --intervals HOUR DAY WEEK MONTH
                        Set all intervals at once
  -r HOURS DAYS WEEKS MONTHS YEARS, --retentions HOURS DAYS WEEKS MONTHS YEARS
                        Set all retentions at once
  --int_hour HOURS      hourly interval for snapshots
  --int_day DAYS        daily interval for snapshots
  --int_week WEEKS      weekly interval for snapshots
  --int_month MONTHS    monthly interval for snapshots
  --ret_hour HOURS      number of hourly snapshots to keep
  --ret_day DAYS        number of daily snapshots to keep
  --ret_week WEEKS      number of weekly snapshots to keep
  --ret_month MONTHS    number of monthly snapshots to keep
  --ret_year YEARS      number of yearly snapshots to keep
  --pre PRE_COMMAND     command to run on host to prepare for starting EBS
                        snapshot (will not be run if volume group is not
                        attached)
  --post POST_COMMAND   command to run on host after snapshot (will not be run
                        if volume group is not attached)
  -d DESCRIPTION, --description DESCRIPTION
                        description to add to snapshot
  -m MOUNT_POINT, --mount-point MOUNT_POINT
                        mount point of the volume group to snapshot

ams snapshot schedule add volume (volume_group_id)

Schedule snapshots for a specific volume_group_id. Least flexible of all the schedule creations, as it will only snapshot the volume group it is assigned to do. If the volume group is no longer in use, snapshots will continue to be created.
int_* settings affect how often a snapshot is performed/retained. eg. 2 for hourly will take an "hourly" snapshot every other hour, 2 for daily will take a "daily" snapshot every other day.
ret_* settings affect how many of each type of snapshot to keep. eg. 24 for hours will keep the last 24 "hourly" snapshots (not necessarily the last 24 hours if "hourly" interval is not 1). Setting the retain value for any of the types to 0 disables that one.
If --intervals or --retentions are set, they will override the single int_* and ret_* arguments
PRE and POST are commands that will be run before and after the snapshot, and provide a means to ensure that data is in a consistent state before snapshotting and revert back to normal operation after snapshot has begun.
Description is written as metadata to the snapshot itself and will show up in the EC2 console.

Required arguments: volume_group_id

Defaults:

  • int_hour 1
  • int_day 1
  • int_week 1
  • int_month 1
  • ret_hour 24
  • ret_day 14
  • ret_week 4
  • ret_month 12
  • ret_year 3

Arguments:

  -i HOUR DAY WEEK MONTH, --intervals HOUR DAY WEEK MONTH
                        Set all intervals at once
  -r HOURS DAYS WEEKS MONTHS YEARS, --retentions HOURS DAYS WEEKS MONTHS YEARS
                        Set all retentions at once
  --int_hour HOURS      hourly interval for snapshots
  --int_day DAYS        daily interval for snapshots
  --int_week WEEKS      weekly interval for snapshots
  --int_month MONTHS    monthly interval for snapshots
  --ret_hour HOURS      number of hourly snapshots to keep
  --ret_day DAYS        number of daily snapshots to keep
  --ret_week WEEKS      number of weekly snapshots to keep
  --ret_month MONTHS    number of monthly snapshots to keep
  --ret_year YEARS      number of yearly snapshots to keep
  --pre PRE_COMMAND     command to run on host to prepare for starting EBS
                        snapshot (will not be run if volume group is not
                        attached)
  --post POST_COMMAND   command to run on host after snapshot (will not be run
                        if volume group is not attached)
  -d DESCRIPTION, --description DESCRIPTION
                        description to add to snapshot

ams snapshot schedule edit (schedule_id)

Edit an existing snapshot schedule by schedule_id.
int_* settings affect how often a snapshot is performed/retained. eg. 2 for hourly will take an "hourly" snapshot every other hour, 2 for daily will take a "daily" snapshot every other day.
ret_* settings affect how many of each type of snapshot to keep. eg. 24 for hours will keep the last 24 "hourly" snapshots (not necessarily the last 24 hours if "hourly" interval is not 1). Setting the retain value for any of the types to 0 disables that one.
If --intervals or --retentions are set, they will override the single int_* and ret_* arguments as well as overwrite all the single settings in the database. If you only want to update a single setting, use the single versions of the arguments
PRE and POST are commands that will be run before and after the snapshot, and provide a means to ensure that data is in a consistent state before snapshotting and revert back to normal operation after snapshot has begun.
Description is written as metadata to the snapshot itself and will show up in the EC2 console. Changing the description does not update the descriptions on snapshots that have already been created; it only changes the description for new snapshots going forward
At this time, changing a snapshot schedule from volume/host/instance type to any other type is not supported. Delete the current schedule and add a new one with different type but the same settings to achieve this functionality.

Required arguments: schedule_id

Arguments:

  -i HOUR DAY WEEK MONTH, --intervals HOUR DAY WEEK MONTH
                        Set all intervals at once
  -r HOURS DAYS WEEKS MONTHS YEARS, --retentions HOURS DAYS WEEKS MONTHS YEARS
                        Set all retentions at once
  --int_hour HOURS      hourly interval for snapshots
  --int_day DAYS        daily interval for snapshots
  --int_week WEEKS      weekly interval for snapshots
  --int_month MONTHS    monthly interval for snapshots
  --ret_hour HOURS      number of hourly snapshots to keep
  --ret_day DAYS        number of daily snapshots to keep
  --ret_week WEEKS      number of weekly snapshots to keep
  --ret_month MONTHS    number of monthly snapshots to keep
  --ret_year YEARS      number of yearly snapshots to keep
  --pre PRE_COMMAND     command to run on host to prepare for starting EBS
                        snapshot (will not be run if volume group is not
                        attached)
  --post POST_COMMAND   command to run on host after snapshot (will not be run
                        if volume group is not attached)
  -d DESCRIPTION, --description DESCRIPTION
                        description to add to snapshot

ams snapshot schedule delete (schedule_id)

Deletes a specific snapshot schedule. Use ams snapshot schedule list to find the schedule_id of a specific schedule

Required arguments: schedule_id

ams snapshot schedule run [schedule_id]

This is intended to be dropped into a cron on a single host every hour with no arguments.
If a schedule_id is provided then the snapshot for the schedule points to will be created immediately regardless of whether it is scheduled (with a best effort to apply the retention rules so the snapshot will eventually be cleaned up). Take note that if a valid expiry time can be calculated the snapshot will be automatically purged per the rules of the schedule. If you want a snapshot that will not expire use ams snapshot create to create a snapshot.

Arguments:

  --purge      delete expired snapshots after running the schedule

Internals

ams internals database install

This will install the database table for an initial install of AMS.

ams internals database upgrade

This should be run every time that the software is updated to ensure that database schema matches the application's expectation.