Allan Jude

Describe a system of using ZFS Boot Environments to quickly, safely, and atomically upgrade 100s of remote machines.

Overview: * How the file system hierarchy is modified to allow the systems to be upgraded in-place * How we use ZFS to create and deploy the boot environments * Simplifying the creation of the BEs using poudriere image * Extending poudriere image to support ZFS * Using zfsbootcfg to boot a new BE once * How we determine if the BE “works” and should be promoted to the default

Motivation: * We often have only SSH access to the machines, so we needed something that could self-recover just by power cycling the machine * No longer need to use freebsd-update, mergemaster, or etcupdate

An in-depth look at how caching works in ZFS, specifically the Adaptive Replacement Cache (ARC) algorithm. Assumes no prior knowledge of ZFS or operating system internals.

ZFS does not use the standard buffer cache provided by the operating system, but instead uses the more advanced "Adaptive Replacement Cache" (ARC).

• What is a cache
• How most caches work (LRU)
• Pros
• Cons
• What makes the ARC different?
• Recently Used
• Frequently Used
• Ghost Lists
• What makes the ARC Adaptive?
• Access Patterns (How the ARC adjusts over time)
• Compressed ARC
• Advantages over compressed memory or swapcache
• Tuning for...
• File Server
• iSCSI Target
• Database
• Hypervisor

How to apply various FreeBSD technologies and features in combination with ZFS to get the most out of a system, be it a Server, Laptop, Embedded, or an Appliance.

ZFS is an advanced filesystem, but that doesn't mean it is complicated or hard to understand.

After an overview of the basics of what ZFS is and how it works, we'll delve into how to integrate ZFS into a system or product. Topics to include: * Boot Environments and how to use them - new boot environment management tool and library from Google Summer of Code - case studies / use cases * Advanced Boot Configuration (zfsbootcfg/nextboot) - Boot the new system image only once - Fall back to previous version if new system image fails to stay up for 5 minutes three consecutive times * Automated Boot Recovery (Appliance style failover/failsafe) - Fall back to rescue image if system fails to boot repeatedly (required for environments like Amazon where there is no console) * GELI Encryption Support - Improvements to allow unattended booting * Appliance Upgrading with ZFS (including zpool checkpoints) - Create a checkpoint before upgrading an appliance, revert to checkpoint if upgrade is not completely successful. Checkpoints can undo things that a snapshot cannot, like renaming or destroying an entire filesystem.

Techniques discussed will be of interest to most everyone, from users and developers for laptops and servers, to appliance manufacturers, and embedded products.

FreeBSD has supported disk encryption with GBDE and GELI since 2002 and 2005 respectively. However, booting the system required storing the loader and kernel unencrypted so that the requisite GEOM module could be loaded to handle decryption. This became a significantly larger stumbling block with the introduction of ZFS, as having multiple separate partitions detracts from the advantages of ZFS, and also causes headaches when upgrading the operating system. With the growing popularity of ZFS Boot Environments, a solution was needed that allowed the kernel and loader to remain part of the primary file system, even if it was encrypted. This paper provides an overview of the design of the GELI enabled BIOS boot code and loader, as well as the numerous challenges encountered during their development.

A walk through the tale of woe that was implementing support for GELI in the FreeBSD bootcode and loader. Hear the story of a very junior developer persisting through countless complications and roadblocks to finally arrive at working code. Learn just how complicated it is to boot a computer, and how much worse it can get. In the end, we are left with working ZFS Boot Environments, even with fully encrypted pools.

A tour through some lesser known features, subcommands, and recipes for doing interesting things with ZFS, presented by the co-author of "FreeBSD Mastery: ZFS"

ZFS is well known for being a rock solid storage system with a plethora of features. Yet, most users are unaware of just how powerful ZFS is. ZFS is so fundamentally different from other file systems that new and interesting things are possible. This presentation will demonstrate a number of very interesting configurations that sysadmins, developers, and others may find very useful.

The presentation starts with a very brief summary of ZFS's features, then immediately delves into interesting hacks. * Script-ability of ZFS * Abusing properties and inheritance * Mount point trickery * Delegation hacks and recipes * Applying limits to users and containers * Using ZFS to juggle development workflows * ZFS replication for conference attendees and road warriors * Lesser known subcommands * Abusing zdb * Storing configuration in ZFS

To be followed by an overview of recent and upcoming developments in OpenZFS, and then questions.