Show filters Hide filters

Refine your search

Publication Year
1-24 out of 24 results
Change view
  • Sort by:
58:02 Berkeley System Distribution (BSD), Andrea Ross English 2014

FreeBSD's Ext2 Implementation

While FreeBSD has been experiencing a huge momentum with ZFS and UFS2 continues to improve in exciting ways, another UFS variant has seen some advances: Ext2fs is trying to catch up with the new times. FreeBSD's ext2 implementation has a huge history: it started it's life as part of BSD-Lites for CMU's Mach. It got ported to FreeBSD and after some rust gathering it eventually got important merges from NetBSD's port so now it is completely GPL free and is supporting new features. Ext2 uses many concepts that are based on UFS but, for good or for bad, Ext2 has never been considered important enough to compete with the local UFS filesystem in the BSDs. Both filesystems have also taken different design decisions over the years and it is not always clear which developments are a clear win on either side. The core of both filesystems is indeed very similar and that similarity has made it relatively easy to adopt in FreeBSD's Ext2fs many enhancements based on FreeBSD's UFS support. Recent years have seen some developments based on successful Google Summer of Code Projects so FreeBSD's driver has support for the directory index and read-only support for extents. This talk presents the recent developments in ext2fs for FreeBSD and attempts to compare the linux ext2/3/4 features with the typical UFS filesystem in the BSDs.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
33:53 Berkeley System Distribution (BSD), Andrea Ross English 2014

Migrating from Linux to FreeBSD as a backend data store

In this talk I would like to discuss the process I went through in replacing a research file server's OS from Linux to FreeBSD and that of its backup system. Initially, the system was hosting a custom 32bit Gentoo Linux instance and was able to host approximately 2.5TB of data over two file systems backed by a RAID-5 software RAID implementation. The solution I implemented allowed me to leverage all the installed disks as one large ZFSv2 data volume by booting the system as a diskless FreeBSD system. I then utilized istgt (an iSCSI target software package) to offer a large disk to one of the group's 64bit Gentoo Linux compute nodes whereby I could then launch autofs and racoon services. I would like to present this material to the BSDCan attendees to share my experience with others and to show that others can accomplish the same task without spending a large amount of money on proprietary software.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
36:18 Berkeley System Distribution (BSD), Andrea Ross English 2014

Keeping Current

Detailed discussion of tips, tools and strategies for maintaining internal software forks with a particular emphasis on FreeBSD and its use in EMC's Isilon Storage Division. Occasionally you need to fork a project. You probably still want to track the upstream version but you've got local modifications you need to maintain. The EMC Isilon Storage Division does this with FreeBSD in a big way. This presentation will go through what happens when you don't keep up to date and how you can get back up to date in a way that can still mesh with your release process. It'll also go through tools you can use to help the process along and keep your project managers happy and strategies for making sure you don't get behind again.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
48:21 Berkeley System Distribution (BSD), Andrea Ross English 2014

FreeBSD Test Suite

FreeBSD has seen the birth of its test suite in 2013 and, while the result is heavily based on experience gained in the NetBSD project, there are significant differences between the two. Come to this talk to learn what the FreeBSD test suite is by looking at the various design choices, the tools in use, the roadmap to the final product and, if you are into coding, some actual examples of tests and the process to plug them into the suite. FreeBSD has been in the need for an automated test suite for a long time. During the last few years, various alternatives have been investigated and, finally, the beginnings of a tangible test suite surfaced in 2013. The FreeBSD test suite is now a reality: the system builds and installs tests out of the box; the test suite can be trivially executed with a single command; the testing cluster is up and running, publishing results for test runs about twice per hour; and there is a lot of ongoing work to make the current setup better. This talk will provide you with a general overview of the new FreeBSD test suite, both from the perspective of an end user and of a developer. We will cover many different areas, including but not limited to: the structure of the test suite, covering the use of Kyua as the run-time engine and the various "frameworks" available to implement test programs; the set up of the continuous testing cluster, to understand how its activity benefits developers; the differences between the FreeBSD and NetBSD test suites, and how the two will eventually converge; the future plans for the test suite, especially in the area of reporting and kernel-level testing; and, to conclude, we will get our hands dirty by seeing actual code for a bunch of simple tests. You will hopefully get a sense of the non-existent difficulty to start writing tests en-masse!
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
44:36 Berkeley System Distribution (BSD), Andrea Ross English 2014

LibreSSL - An OpenSSL replacement

LibreSSL is a recently started initiative by the OpenBSD project to fork and clean up the OpenSSL code base. We have been working on this rather intensively for about a month now. I will be discussing the Origins of LibreSSL (The OpenBSD Fork of OpenSSL) - including why we decided to fork, what we are hoping to achieve, as well as a number of examples of the sorts of code changes we are doing, and the sorts of issues we have found in the code base and how we are changing things. All who are potentially interested in something better than OpenSSL may find this talk of interest.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
36:39 Berkeley System Distribution (BSD), Andrea Ross English 2014

Transparent Superpages for FreeBSD on ARM

The talk covers recent work on providing transparent superpages support for the FreeBSD operating system on ARM. The concept of superpages is a virtual memory system optimization, which allow for efficient use of the TLB translations, effectively reducing the overhead related to the memory management. This technique can significantly increase system's performance at the interface between CPU and main memory thus affecting its overall efficiency. The lecture focuses on presenting superpages mechanism supported by the FreeBSD in the context of its implementation for the ARM architecture. Principles of the virtual memory operations are briefly presented to illustrate the requirements and limitations standing before the related subsystem. The talk provides description of the virtual memory system architecture introduced in ARMv6/v7 compliant processors and the opportunities to take advantage of the superpages technique. The primary focus of the presentation is to elaborate on how the superpages functionality was implemented on FreeBSD/arm and what are the results of its application. The talk presents real-life measurements and benchmarks performed on modern, multiprocessor ARM platforms. Hence, the actual achievements and areas of application are shown. Finally, the article summarizes the integration process to the mainline FreeBSD and discusses the areas of future work and improvements.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
54:41 Berkeley System Distribution (BSD), Andrea Ross English 2014

Using BGP for realtime import and export of spam whitelist/blacklist entries

The results of using BGP for realtime import and export of spam whitelist/blacklist entries In early 2013, I introduced a new method to distribute spam whitelist/blacklist entries. Now, I am able to present the results of 1 year of usage. In the battle against Spam, many mail server admins collect and distribute IP addresses of systems that send them Spam. However, distribution of these lists are traditionally limited to 2 methods. is periodically downloading this list from a source, usually a web server often causing massive load and slowness at the top of the hour. is a real-time lookup against an external provider (such as dns-rbls) so your response time is dependent on how fast they respond or timeout. This talk discusses a 3rd solution: using BGP to distribute the IP addresses in a real-time manner.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
38:36 Berkeley System Distribution (BSD), Andrea Ross English 2014

The architecture of the new solver in pkg

In the context of package management, the solver is an algorithm (or set of algorithms) to resolve dependencies and conflicts. The solver must handle options, upgrades, multiple repos, locally installed software, as well as other factors. The upcoming 1.3 release of pkg will have the new solver that has some important consequences. This talk is dedicated to the design concepts of the new solver in pkg management system (pkg-ng initially). In this talk, I describe the basic architecture of the solver, ideas used and the consequences of using this algorithm. Moreover, this talk describes the proposed pkg and ports architecture to simplify binary packages and ports using for all FreeBSD users. The proposed talk is oriented to the wide auditory of FreeBSD users and describes the architectural design of pkg and ports that is going to be implemented. I concentrate on the following topics: the interaction of ports and packages how pkg may be used for ports management what are the alternatives used by other packages management systems (yum/apt/zypper/MacPorts) the basic description of the SAT problem and how this algorithm can be applied to packages management multiple repos handling the improved support of custom options advanced conflicts and dependencies fine-grained packages distribution base system interaction the speed estimations This talk is intended to improve the understanding of pkg internals and the future plans of FreeBSD packages and ports development.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
30:46 Berkeley System Distribution (BSD), Andrea Ross English 2014

BSDCan 2014 Keynote, FlightAware

  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
1:11:28 Berkeley System Distribution (BSD), Andrea Ross English 2014

New trends in mandoc

Since its first presentation at BSDCan 2011, the mandoc(1) documentation formatting system has gained a lot of traction in all four major BSD operating systems and beyond. This talk will provide an update regarding recent feature additions, in particular the mdoc(7) to man(7) converter to produce high quality manual formatting for portable software packages, and the new SQLite3-based mandocdb(8) utility, a drop-in replacement for the traditional makewhatis(8)/apropos(1) combo, providing markup-sensitive search capabilities and flexible search result output formats. I will give an overview of the adoption status of mandoc(1) and mandocdb(8) in OpenBSD, DragonFly, NetBSD, and FreeBSD as well as a few other free operating systems, comment on mandoc versus groff usage in the OpenBSD ports tree including recommendations for handling manuals in packaging systems in general, and comment on possible future development directions.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
57:37 Berkeley System Distribution (BSD), Andrea Ross English 2014

Shellscripts and Commands

Shellscripts and Commands: practical development method from large-scale academic research to mission-critical business system Universal Shell Programming Liboratory has developed many kinds of mission-critical business systems for many companies based on their own unique and innovative development method for SI and big-data processing. The main idea of that might be too simple to believe in: text files, commands and shellscripts. You can convert almost any kind of your data into text. Then the development method tells you how to organize your data in what directories. As a result, it realizes fast processing and easy coding. This method is applicable in many areas, from as small tools to larg-scale data processing. I describe their development method in a tangible way using Open usp Tukubai and FreeBSD.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
45:18 Berkeley System Distribution (BSD), Andrea Ross English 2014

Introducing ASLR in FreeBSD

Address-space layout randomization (ASLR) has existed in many operating systems for a number of years. The most famous implementation is the PaX patch for Linux's kernel. This presentation introduces and announces an ASLR implementation based on PaX for FreeBSD/amd64. Details regarding how ASLR has been ported to FreeBSD and some advanced features will be presented. FreeBSD will soon be getting a port of PaX to 11-CURRENT/amd64. This presentation details changes to how ELF executables are loaded in memory and innovative workarounds for legacy applications that don't support (or misbehave) ASLR. Jails can have their own ASLR settings. Misbehaving applications can be run in a jail with ASLR turned off, while ASLR remains turned on in the other jails and in the host.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
1:11:39 Berkeley System Distribution (BSD), Andrea Ross English 2014

FreeBSD bmake and meta mode

The Junos "meta mode" build is the coolest thing since sliced bread. The FreeBSD projects/bmake branch provides a proof of concept for the more general use of this technology. The latest release of FreeBSD uses bmake by default, a prerequisite for the "meta mode" build. This talk will cover the adoption of bmake and how the projects/bmake branch differs from traditional FreeBSD build.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
39:21 Berkeley System Distribution (BSD), Andrea Ross English 2014

Taking the red pill

The Xen hypervisor is a Open Source Type 1 hypervisor, it's widely used on production environments like Amazon EC2 and Rackspace. Since it's inception, one of the focuses of Xen was to be an OS agnostic hypervisor, allowing any kind of OS (with proper Xen support) to act as DomU/Dom0. This talk will cover how the Xen community works, together with an explanation of the ongoing work in FreeBSD in order to improve Xen support. This talk will cover the following points: Basic Xen description and specific Xen concepts. How the Xen community works (compared to BSD communities). A look into new Xen features (PVH). Work being done in FreeBSD improving Xen support. Probably a small demo to highlight Xen features.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
1:04:51 Berkeley System Distribution (BSD), Andrea Ross English 2014

Userspace Networking with libuinet

libuinet is a userspace library version of the FreeBSD TCP/IP stack that also includes extensions to the base stack functionality that make it particularly useful in network infrastructure equipment. This talk will cover its design goals, implementation, current and potential uses, and performance. libuinet was originally conceived as a way to bring highly scalable transparent proxy functionality to the free, portable TCP proxy WANProxy (http://wanproxy.org). To this end, libuinet extends the base FreeBSD TCP/IP stack feature set to include 'promiscuous sockets', which allow listens to capture connection attempts across VLANs (including nested), any IP address, and any port, admit/ignore those attempts based on an application-supplied filter, and retrieve the complete L2 and L3 details of admitted connections. Promiscuous socket functionality also allows active connections to fully specify their L2 and L3 identity. In this mode, libuinet has been shown to scale to 1 million active connections concurrent with 1 million listen sockets, with those million connections distributed in multiple ways across the VLAN and 4-tuple TCP addressing space. Implementation of another extension to the stack, 'passive sockets', is currently underway and targeted for completion by the end of 1Q2014. Passive sockets provide for reassembly of both data streams in a TCP connection, along with a missing-frame notification mechanism, based on a copy of the packet stream flowing between the connection endpoints (e.g., via a SPAN port).
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
57:28 Berkeley System Distribution (BSD), Andrea Ross English 2014

BSD/ARM Kernel Internals

In this talk, I'll discuss how BSD kernels interface with the ARM processor. I will cover the kernel internals of the FreeBSD and NetBSD ARM ports, focusing on ARMv7 primarily. I will discuss how booting, memory management, exceptions, and interrupts work using plenty of BSD code. This talk is meant to be a quick start guide for BSD hackers who aren't familiar with the ARM architecture.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
39:17 Berkeley System Distribution (BSD), Andrea Ross English 2014

IPv6 Transitioning mechanisms on the BSDs

The growth pace of IPv6 adoption is still slow, but constantly increasing, as more providers and networks migrate. There is one aspect of the adoption that is still underestimated, and it's transition mechanisms, enabling networks speaking different protocols to talk to each other. In may 2013, Switzerland jumped on top of IPv6 utilisation in the world by just having its incumbent operator enable one of these for a large base of its users. This talk will first introduce a handful of different transitioning mechanisms in use, picking the most widely used amongst the plethora of ones available. In the second part, a live demonstration will show the audience how to set up some of them using native tools on OpenBSD and FreeBSD.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
40:20 Berkeley System Distribution (BSD), Andrea Ross English 2014

In-kernel OpenvSwitch on FreeBSD

OpenvSwitch (OVS) is a very popular component for experimenting with Software Defined Networking. OVS is especially useful for interconnecting virtual machines, as it eases VM migration and reconfiguration of routing in the hosting provider. In this paper we will discuss how we ported the in-kernel OVS dataplane to FreeBSD, using the original linux code with minimal modifications and suitable wrappers to build a FreeBSD kernel module, parse netlink messages and convert skbufs and mbufs.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
54:25 Berkeley System Distribution (BSD), Andrea Ross English 2014

Optimizing GELI Performance

Features, like encryption, need to have minimal overhead for them to be widely adopted. If the performance is to slow, few people will use it. The first iteration of AES-XTS using AES-NI in FreeBSD was not much faster than the software version of it. The talk will describe why the AES-XTS algorithm was slow and what was done to improve it. It will cover topics from intrinsics, adding them to gcc and advantages of using them over assembly to how to use HWPC that are included in most modern processors to evaluate performance to identify performance bottle necks. Optimizing code first starts with measuring the performance, but it also requires you to understand the parts of the system so that you can decide if increasing performance is possible. It will cover: 1) Cipher modes and their performance impact 2) Processor performance, understanding pipelining, throughput and latency 3) Other SSE instructions used for XTS tweak factor calculations 4) Intrinsics and their use with GCC and CLANG 5) Using pmcstat and kcachegrind for understand performance. 6) Possible future work to increase the performance beyond what it is today.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
52:52 Berkeley System Distribution (BSD), Andrea Ross English 2014

Capsicum and Casper - more than a lipstick on a pig

Capsicum and Casper are FreeBSD proposal for a clean, robust and intuitive application compartmentalization. Today's sandboxing techniques build on top of existing technologies that weren't really designed for this sort of protection (like chroot(2), rlimit(2), setuid(2), Mandantory Access Control, etc.). Capsicum and Casper provide rich infrastructure for breaking applications into multiple useful sandboxes and thus significantly reducing Trusted Computing Base. Capsicum is a lightweight OS capability and sandbox framework implementing a hybrid capability system model. The Casper daemon enables sandboxed application to use functionality normally unavailable in capability-mode sandboxes. The talk will discuss Capsicum framework, Casper daemon and its services. It will provide introduction based on already implemented examples to those new FreeBSD features. The talk will also present existing portable sandboxing implementations to give clear picture how hacky those solutions are.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
1:03:46 Berkeley System Distribution (BSD), Andrea Ross English 2014

OpenBGPD turns 10 years

The Border Gateway Protocol, BGP, is used on the internet between ISPs to announce reachability of networks. Routers build their routing tables using this information. The global IPv4 routing table has about 470000 entries today. In 2004, I was upset enough with the implementation we were using back then, zebra, to start writing an own one. After showing an early prototype other developers jumped in and helped. Quickly thereafter we had a working BGP implementation that not only I have used ever since then. We'll look at OpenBGPD's design and how it differs from other implementations, the frameworks established and later used for other purposes, and the lessons we learned over the last 10 years.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
50:02 Berkeley System Distribution (BSD), Andrea Ross English 2014

NAND Flash and FreeBSD

An overview of NAND technology, its relevance to FreeBSD, and a discussion of the advantages and disadvantages of the NANDFS and other nand technology in FreeBSD. The first part of this talk will be generally about NAND flash, and is applicable to embedded, PCIe-based flash storage and SSD. The second part of the talk focuses on those technologies in FreeBSD to talk to raw NAND, what can be deployed today, and what additional work will be necessary in the future. NAND Flash technology continues to be important to some market segments in which FreeBSD is deployed. This talk will be in two parts. The first part of the talk will discuss NAND flash generally. The second part of the talk will focus on NAND in a FreeBSD context. In the first part, an overview of AND flash will be presented. Basics about NAND generations, NAND storage strategies (SLC, MLC, TLC), characteristics of NAND, and operational considerations will be discussed. An overview of the trends in NAND evolution and their implications will be presented. The talk will explore how these features apply to SSDs and PCIe cards. This part of the talk will be of interest to anybody using these technologies either directly attached, or in a storage product. The second part of the talk will be more focused on embedding FreeBSD, running FreeBSD from NAND and an exploration of the currently supported technology. Kernel configuration, build parameter and other details necessary to build a system will be presented. Creating a system image is a bigger challenge with NAND flash, because blocks on the part wear out and need to be retired (some blocks come from the factory this way), and how to compensate. The additional features need to be developed to deploy the latest NAND parts may be discussed. The second part of the talk will be more focused more on FreeBSD embedded systems using NAND.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
59:36 Berkeley System Distribution (BSD), Andrea Ross English 2014

OpenZFS: upcoming features and performance enhancements

The OpenZFS project provides a common development hub for all platforms working with open source ZFS code. Currently, it is easy to pull changes from Illumos into FreeBSD, but it is more difficult to submit changes from FreeBSD to Illumos. This talk will discuss how OpenZFS will enable ZFS code and ideas to flow easily between the Illumos, FreeBSD and ZFS on Linux communities. In addition, I will present several important features and performance enhancements to ZFS in FreeBSD, and also discuss forthcoming enhancements that are in the planning phase.
  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
40:12 Berkeley System Distribution (BSD), Andrea Ross English 2014

MIPS router hacking

  • Published: 2014
  • Publisher: Berkeley System Distribution (BSD), Andrea Ross
  • Language: English
out of 1 pages
Loading...
Feedback

Timings

  187 ms - page object
  109 ms - search
    5 ms - highlighting
    2 ms - highlighting/15330
    3 ms - highlighting/15362
    3 ms - highlighting/15352
    3 ms - highlighting/15333
    4 ms - highlighting/15361
    3 ms - highlighting/15365
    3 ms - highlighting/15337
    4 ms - highlighting/15355
    4 ms - highlighting/15350
    4 ms - highlighting/15357
    3 ms - highlighting/15341
    4 ms - highlighting/15363
    2 ms - highlighting/15343
    2 ms - highlighting/15356
    4 ms - highlighting/15353
    2 ms - highlighting/15332
    2 ms - highlighting/15364
    2 ms - highlighting/15349
    3 ms - highlighting/15336
    3 ms - highlighting/15347
    1 ms - highlighting/15344
    1 ms - highlighting/15360
    1 ms - highlighting/15342
    1 ms - highlighting/15366

Version

AV-Portal 3.7.0 (943df4b4639bec127ddc6b93adb0c7d8d995f77c)