Episode Details

This week, we are going to be talking with Aaron Poffenberger, who has much to share about his first-hand experience in infiltrating Linux conferences with BSD-goodness.

This episode was brought to you by

Headlines

Alexander Motin implements CTL High Availability

• CTL HA allows two .head. nodes to be connected to the same set of disks, safely
• An HA storage appliance usually consists of 2 totally separate servers, connected to a shared set of disks in separate JBOD sleds
• The problem with this setup is that if both machines try to use the disks at the same time, bad things will happen
• With CTL HA, the two nodes can communicate, in this case over a special TCP protocol, to coordinate and make sure they do not step on each others toes, allowing safe operation
• The CTL HA implementation in FreeBSD can operate in the following four modes:
• Active/Unavailable -- without interlink between nodes
  • Active/Standby -- with the second node handling only basic LUN discovery and reservation, synchronizing with the first node through the interlink
  • Active/Active -- with both nodes processing commands and accessing the backing storage, synchronizing with the first node through the interlink
  • Active/Proxy -- with second node working as proxy, transferring all commands to the first node for execution through the interlink
• The custom TCP protocol has no authentication, so it should never be enabled on public interfaces
• Doc Update ***

Panel Self-Refresh support lands in DragonFlyBSD

• In what seems almost weekly improvements being made to the Xorg stack for DragonFly, we now have Panel Self-Refresh landing, thanks to Imre Vadász
• Understanding Panel Self-Refresh and More about Panel Self-Refresh
• In a nutshell, the above articles talks about how in the case of static images on the screen, power-savings can be obtained by refreshing static images from display memory (frame-buffer), disabling the video processing of the CPU/GPU and associated pipeline during the process.
• And just for good measure, Imre also committed some further Intel driver cleanup, reducing the diff with Linux 3.17 ***

Introducing Sluice, a new ZFS snapshot management tool

• A new ZFS snapshot management tool written in Python and modeled after Apple.s Time Machine
• Simple command line interface
• No configuration files, settings are stored as ZFS user properties
• Includes simple remote replication support
• Can operate on remote systems with the zfs://user@host/path@snapname url schema
• Future feature list includes .import. command to moved files from non-ZFS storage to ZFS and create a snap