v0.5

New Features

API Server

Plunder now has a native API-Server that makes it configurable via a rest API, this is already exploited through the use of tooling such as pldrctl and Cluster-API. It uses certificates to encrypt communication between a client and the server, these certificates can be generated through the plunder config apiserver <server/client> command

Native ISO support

Plunder can now read directly from an ISO file, meaning that mounting an ISO to local host is no longer required, the ISO can be added into the server configuration:

  - cmdline: ""
    configName: preseed
    initrdPath: ubuntu/install/netboot/ubuntu-installer/amd64/initrd.gz
    isoPath: /nfs/Operating Systems/Ubuntu/ubuntu-16.04.5-server-amd64.iso
    isoPrefix: ubuntu
    kernelPath: ubuntu/install/netboot/ubuntu-installer/amd64/linux

Alternatively, it can be added through the API or with pldrctl

Unleased server watching

Any server that isn't currently being deployed will become part of an unleased group, that can be managed with the --defaultBoot option. This gives the option of forcing a server into a reboot loop until it is required for provisioning.

Tons of bug fixes

• Logging improvements, including an internal logging and streaming package
• Move to GO Modules
• ESXi support
• Fixes to DHCP where machines restarting at the same time could hit a race condition

Additional fixes can be read through by looking at the commit delta.

plndr.io

There is finally an external URL for plunder https://plndr.io !

Plunder v0.4.5

New Features

• The parlay engine will now process JSON and YAML automatically, the use of YAML greatly reduces the size of a deployment map and makes it more human readable
• Extensions to the command action so that data can be send over a pipe to a remotely running process. An example of a use case is doing a docker export locally and piping the image to a remotely running docker import - command
• The parlay engine can now take the remote host authentication details from a local deployment configuration, but now also from a HTTP deployment endpoint.
• The parlay engine can also have it's configuration overwritten through additional flags that can be passed at runtime
• Plunder no longer needs to generate a AA:BB:CC:DD:EE:FF.ipxe file for every host that it will bootstrap, these urls (and file content) are built in memory either when plunder starts or when the http://x.x.x.x/depoyment is updated with a POST.

Fixes

• Various error messages have been fixed
• Moved from sirupsen to Sirupsen for the logging library
• Ensures that if swap is disabled it's also removed in /etc/fstab on Ubuntu 18.04
• Fixes a panic if a ~/plugin directory doesn't exist

Documentation

... all of the documentation has been updated, including new pages around actions and the ui.

Plunder v0.4

A new UI command has been added to the automate ssh section.. This will ultimately be dropped to below ssh once the VMware integration is worked on. However it provides the capability of navigating through a full map of multiple deployments, that have various action on various hosts.

Through the UI a user can select a specific deployment, which actions should be performed and on what hosts.. This makes it very simple to add new hosts or add a missing package/image to a remote host. Further functionality will be added at a later date (such as a UI editor)..

Also a Docker plugin has been started on, so far the functionality is only around management of images between remote hosts.

Demo of the UI feature

Plunder v0.3

There have been numerous changes to the usage of plunder but the documentation should have been updated to reflect that.

Changes:

The following changes have been made:

• The --deployment flag is now --config
• The --deployment flag now will point to a specific deployment in a map
• The --action flag can be used to point to a specific action in a deployment
• The --host flag will point to a specific host in the deployment
  and finally
• The --resume will determine if to continue executing all remaining actions
• Kubeadm support for deploying Kubernetes with plunder
• Plugin model, will be span out to a separate project at a later date
• Various other fixes to logging
• Parallel as well as sequential deployments.

Plunder v0.2

A ton of new functionality has been added:

PXE Booting

• The intial iPXE configuration will pass the MAC address of the boot adapter instead of a device name, this means that the boot process should succeed with multiple adapters
• The deployment configuraton takes an adapter configuration in both global and deployment for a server to ensure that the correct adapter is selected for the next IP address configuration.
• The disk format has been updated to a more modern GPT format, taken from preseed examples that have been written in the last two decades.
• DHCP has been updated to simply not respond to an unknown MAC address, however reboot is still an option when a server doesn't reboot by default.

Remote access

• SSH functionality to run remote commands
• SFTP for upload and Download of files

Automation
Creation of a processing engine called parlay (More captain references .. shoot me)

• Takes the ssh configuration from the deployment.json to reduce the amount of configuration files and details that need passing back and forth.
• Takes JSON files that describe the commands that will be executed on the remote hosts..

TODO:
A LOT of documentation needs doing and probably a move to a /docs folder.

Initial release of plunder

This is the first release of plunder, it is still far from production ready however is stable enough in most deployments. There will be upcoming work for fixes for output in the tftp server proportion as it's overly chatty in default logging mode.