Feature #1974 Virtualbox Can't Be Installed Inside a Jail Due to Kernel Module Dependency 12/29/2012 04:30 PM - Marcus Ahlberg

Total Page:16

File Type:pdf, Size:1020Kb

Feature #1974 Virtualbox Can't Be Installed Inside a Jail Due to Kernel Module Dependency 12/29/2012 04:30 PM - Marcus Ahlberg FreeNAS - Feature #1974 VirtualBox can't be installed inside a Jail due to kernel module dependency 12/29/2012 04:30 PM - Marcus Ahlberg Status: Resolved Estimated time: 0.00 hour Priority: Nice to have Assignee: John Hixson Category: Target version: 9.2.1.6-RELEASE Severity: New Needs Merging: Yes Reason for Closing: Needs Automation: No Reason for Blocked: Support Suite Ticket: n/a Needs QA: Yes Hardware Configuration: Needs Doc: Yes Description I have been trying to find a solution on how to install [[VirtualBox]] on a [[FreeNAS]] host. With memory so cheap, the power of a [[FreeNAS]] box should be enough for a couple of virtual servers. The problem I have come across is the fact that [[FreeBSD]] jails share the kernel with the host and Virtualbox needs to load a module into the kernel. I have found a possible solution for installing the kernel module on the host and then installing virtualbox inside a jail. This solution however assumes a full [[FreeBSD]] host where the ports tree can be used. http://forums.freebsd.org/showthread.php?t=21897 I quickly tried to install virtualbox-ose-kmod-legacy using pkg_add on the host, but I couldn't load it due to "KLD vboxnetflt.ko: depends on netgraph - not available or version missmatch". I haven't dug any deeper into this problem. What I call for is support for software add-ons that require kernel modules to be loaded on the host. History #1 - 12/29/2012 04:57 PM - Josh Paetzel The first stopper you are hitting here is the lack of the netgraph kernel module on [[FreeNAS]]. I'll think about this some, there are lots of hidden gotchyas. Perhaps the best solution is shipping with the vbox modules, that seems like a reasonable thing to me. #2 - 12/29/2012 05:27 PM - Josh Paetzel I've looked at this a bit more indepth. Running virtualbox in a jail is a unique case, but it's very useful functionality. What I'm going to do is set [[FreeNAS]] up so this just works. #3 - 04/09/2013 04:01 PM - jlohiser - Is this still in the works or is this issue dead? I would like to run a Linux + Asterisk server as a guest on top of [[FreeNAS]]. #4 - 05/31/2013 03:17 PM - Marcus Ahlberg I would also like an update on this. #5 - 06/02/2013 05:26 PM - 8ender - I'd also love to see this as well. I tried setting up a proxmox server and running Freenas under that but its just not as good as Freenas bare metal. I have lots of spare cycles on this server and it makes so much more sense to have Freenas run bare metal and then have a Virtualbox jail to spawn 10/01/2021 1/9 other servers with, using the freenas mounts for storage. #6 - 10/11/2013 03:14 PM - Patrick Private I'm running VirtualBox inside 9.x jail (also did that in 8.x) but lacking netgraph brings lot of issues. You can't use bridged mode for your VM's - and that's a big deal. Missing netgraph module that is needed as dependency for native VB modules prevents from using bridged mode interfaces in VM's and forces to use NAT on jailed bridge interface. This is sub-optimal solution with multiple limitations to "FreeNass + Virtualization" as whole. Simply adding/ticking netgraph kernel compile option, would helped lot of us - that wants to stick to single-box solution that FN is perfectly suited to + using jails for simpler tasks and VM's for home lab/more advanced scenarios. The obvious reason for FreeNas to support VB is that, sometimes powerful boxes that runs FreeNas could be easily re-used to do some visualization as well. Adding internal storage mapping that eliminate need for NFS/iSCSI makes FreeNas perfect single box, green IT solution for SMB/home. #7 - 10/11/2013 05:03 PM - Josh Paetzel Testing a build with netgraph enabled in the kernel. #8 - 11/02/2013 12:57 AM - Albert Marin Hi Patrick, could you post the procedure to implement VirtualBox inside 9.x jail please? #9 - 11/13/2013 07:45 PM - Marcus Ahlberg Patrick, I would love to hear about your experiences. With or without netgraph. #10 - 11/13/2013 10:48 PM - Patrick Private Albert, Marcus, Recently downloaded newest 9.2 alpha and there are some good news and some bad :D Good news: on host system VB network driver needed for bridged mode loads without any complains. That's major step forward. The bad news is that guest starting from jail cause kernel dump (.!.). I've checked some options around but it was quite hopeless [multiple dumps :D]. Now I need to do testing on real hardware as I was testing VB inside guest in jail on FreeNas host that was running inside VB guest on my Windows. That can be a reason for my failure. So next tests will be on physical PC. @Albert: I've applied multiple hacks in addition to guides available on web (there are some good howtos) to make is running for me. There are two major steps not covered in any guide that is: - allow.sysvipc=1 you should put on jail meta "jail.flags" file not on host directly as guides show - doing NAT'ed VB you need to go 'deep' into VB command line guides to set up proper port forwarding in order to be able access your virtual PC. (I did it once and unfortunately I can't recall exact settings now) I'll post on progress once done. #11 - 11/17/2013 11:23 PM - Jordan Hubbard - Category set to 18 - Status changed from Unscreened to Screened - Target version set to 49 10/01/2021 2/9 #12 - 01/02/2014 08:42 PM - Patrick Private Found some time to check it against 9.2 stable. As it worked before in 9.2 RC build (but only without bridged mode in VB guest) now whole VB kernel support seems to be broken 100%. Where you could load vboxdrv.ko, it's no more possible. Looks like already reported here: https://bugs.freenas.org/issues/3701. For those using VB you just need to roll-back to 9.1 :(. #13 - 01/05/2014 01:52 PM - Patrick Private Update: kernel modules provided by Josh Paetzel here: https://bugs.freenas.org/issues/3701 works and indeed system crash issue is resolved now (still not in stock 9.2). But as pointed out later in #3701 bridged adapters doesn't work. On interface you can see: - in guest: some (strange but not all?) outgoing arps - at host on bridged IF: arp requests WITH arp replies :), -- also legitymate outgoing packets like ICMP queries but without replies What is interesting while trying to do ngctl to check hooks etc. you get "ngctl: can't create node: No such file or directory". In google you get only ARM related topics related to such error so I guess it's some kind of missing dependency. #14 - 01/09/2014 01:01 AM - Ian Pitcher I've seen similar network traffic with the drivers compiled with the VIMAGE patch/option, but not crashing is good! I got ngctl to run by downloading the kernel source at https://github.com/trueos/trueos/releases/tag/9.2.0-RELEASE and replacing /usr/src in my portjail with its contents (this also helps if you want to compile your own vbox modules). I determined that ngctl needed ng_socket.ko loaded, so I did: Within port jail: cd /usr/src/sys/modules/netgraph make; make install Exit the jail and loaded the module on the host: kldload /[path to root of jail]/boot/kernel/ng_socket.ko Re-enter jail, start VM with bridged NIC, and check ngctl output: [root@box /]# ngctl list There are 2 total nodes: Name: vboxnetflt_epair1b Type: vboxnetflt ID: 00000003 Num hooks: 0 Name: ngctl29256 Type: socket ID: 00000004 Num hooks: 0 I don't know anything about netgraph, so I don't know what I should expect to see there. I did compile if_tap.ko and get bridged networking going with that, but it has a major downside: whenever the application accessing the tunnel device quits (ie. VirtualBox), it crashes the kernel. I suspect it's also VIMAGE related, as I saw a PR describing the same behavior here: http://www.freebsd.org/cgi/query-pr.cgi?pr=158686 -- perhaps this is a regression? And since I'm throwing whatever garbage I can at the wall to see what sticks (sorry), there's also this: http://lists.freebsd.org/pipermail/freebsd-questions/2013-May/251160.html I did try to load ng_ether.ko and got "link_elf_obj: symbol ifnet undefined, linker_load_file: Unsupported file type" so maybe the kernel needs to be complied with options NETGRAPH_ETHER 10/01/2021 3/9 #15 - 01/15/2014 07:37 PM - Ian Pitcher - File ng_ether.ko added - File ng_socket.ko added After much horsing around, I can say for certain now that NETGRAPH_ETHER (or ng_ether.ko) are necessary for this to work, and it works with or without VIMAGE enabled for the jail. Here's the output from ngctl in the jail while VBoxHeadless is running a VM: [root@box /]# ngctl list There are 3 total nodes: Name: epair0b Type: ether ID: 00000001 Num hooks: 2 Name: vboxnetflt_epair0b Type: vboxnetflt ID: 00000002 Num hooks: 2 Name: ngctl6630 Type: socket ID: 00000006 Num hooks: 0 For good measure, here's what it looks like on the host itself: antarctica# ngctl list There are 6 total nodes: Name: re0 Type: ether ID: 00000001 Num hooks: 0 Name: ipfw0 Type: ether ID: 00000002 Num hooks: 0 Name: bridge0 Type: ether ID: 00000003 Num hooks: 0 Name: epair0a Type: ether ID: 00000004 Num hooks: 0 Name: epair1a Type: ether ID: 00000006 Num hooks: 0 Name: ngctl6632 Type: socket ID: 0000000c Num hooks: 0 In case anyone else would like to test, I'm attaching my ng_socket.ko and ng_ether.ko modules for FreeNAS 9.2.0-RELEASE here.
Recommended publications
  • The Challenges of Dynamic Network Interfaces
    The Challenges of Dynamic Network Interfaces by Brooks Davis brooks@{aero,FreeBSD}.org The Aerospace Corporation The FreeBSD Project EuroBSDCon 2004 October 29-31, 2004 Karlsruhe, Germany Introduction ● History of Dynamic Interfaces ● Problems ● Possible Solutions ● Advice to Implementors ● Future Work Early UNIX ● Totally static. ● All devices must be compiled in to kernel ● Fast and easy to program ● Difficult to maintain as the number of devices grows Autoconfiguration ● Introduced in 4.1BSD (June 1981) ● One kernel can serve multiple hardware configurations ● Probe – Test for existence of devices, either using stored addresses or matching devices on self-identifying buses ● Attach – Allocate a driver instance (as of 6.0, this must be fully dynamic) Kernel Modules ● Allows drivers to be added at run time ● LKM (Loadable Kernel Modules) – Introduced in 2.0 by Terry Lambert – Modeled after the facility in SunOS ● KLD (dynamic kernel linker) – Introduced along with newbus in 3.0 by Doug Rabson – Added a generic if_detach() function PC Card & CardBus ● Initial PC Card (PCMCIA) support via PAO in 2.0 ● Fairly good support in 3.0 ● Most PAO changes merged in 4.0 – PAO development ceased ● CardBus support in 5.0 Other Removable Devices ● USB Ethernet (4.0) ● Firewire (fwe(4) in 4.8, fwip(4) in 5.3) ● Bluetooth (5.2) ● Hot plug PCI ● Compact PCI ● PCI Express ● Express Card Netgraph ● Node implement network functions ● Arbitrary connection of nodes allowed ● ng_iface(4) node creates interfaces on demand Interface Cloning ● Handles most pseudo
    [Show full text]
  • Katalog Elektronskih Knjiga
    KATALOG ELEKTRONSKIH KNJIGA Br Autor Naziv Godina ISBN Str. Porijeklo izdavanja 1 Peter Kent Pay Per Click Search 2006 0-471-74594-3 130 Kupovina Engine Marketing for Dummies 2 Terry Large Access 1 2007 Internet Freeware 3 Kevin Smith Excel Lassons & Tutorials 2004 Internet Freeware 4 Terry Michael Photografy Tutorials 2006 Internet Freeware Janine Peterson Phil Pivnick 5 Jake Ludington Converting Vinyl LPs 2003 Internet Freeware to CD 6 Allen Wyatt Cleaning Windows XP 2004 0-7645-7311-X Poklon for Dummies 7 Peter Kent Sarch Engine Optimization 2006 0-4717-5441-2 Kupovina for Dummies 8 Terry Large Access 2 2007 Internet Freeware 9 Dirk Dupon How to write, create, 2005 Internet Freeware promote and sell E-books on the Internet 10 Chayden Bates eBook Marketing 2000 Internet Freeware Explained 11 Kevin Sinclair How To Choose A 1999 Internet Freeware Homebased Bussines 12 Bob McElwain 101 Newbie-Frendly Tips 2001 Internet Freeware 13 Windows Basics 2004 Poklon 14 Michael Abrash Zen of Graphic 2005 Poklon Programming, 2. izdanje 15 13 Hot Internet 2000 Internet Freeware Moneymaking Methods 16 K. Williams The Complete HTML 1998 Poklon Teacher 17 C. Darwin On the Origin of Species Internet Freeware 2/175 Br Autor Naziv Godina ISBN Str. Porijeklo izdavanja 18 C. Darwin The Variation of Animals Internet Freeware 19 Bruce Eckel Thinking in C++, Vol 1 2000 Internet Freeware 20 Bruce Eckel Thinking in C++, Vol 2 2000 Internet Freeware 21 James Parton Captains of Industry 1890 399 Internet Freeware 22 Bruno R. Preiss Data Structures and 1998 Internet
    [Show full text]
  • Ebook - Informations About Operating Systems Version: August 15, 2006 | Download
    eBook - Informations about Operating Systems Version: August 15, 2006 | Download: www.operating-system.org AIX Internet: AIX AmigaOS Internet: AmigaOS AtheOS Internet: AtheOS BeIA Internet: BeIA BeOS Internet: BeOS BSDi Internet: BSDi CP/M Internet: CP/M Darwin Internet: Darwin EPOC Internet: EPOC FreeBSD Internet: FreeBSD HP-UX Internet: HP-UX Hurd Internet: Hurd Inferno Internet: Inferno IRIX Internet: IRIX JavaOS Internet: JavaOS LFS Internet: LFS Linspire Internet: Linspire Linux Internet: Linux MacOS Internet: MacOS Minix Internet: Minix MorphOS Internet: MorphOS MS-DOS Internet: MS-DOS MVS Internet: MVS NetBSD Internet: NetBSD NetWare Internet: NetWare Newdeal Internet: Newdeal NEXTSTEP Internet: NEXTSTEP OpenBSD Internet: OpenBSD OS/2 Internet: OS/2 Further operating systems Internet: Further operating systems PalmOS Internet: PalmOS Plan9 Internet: Plan9 QNX Internet: QNX RiscOS Internet: RiscOS Solaris Internet: Solaris SuSE Linux Internet: SuSE Linux Unicos Internet: Unicos Unix Internet: Unix Unixware Internet: Unixware Windows 2000 Internet: Windows 2000 Windows 3.11 Internet: Windows 3.11 Windows 95 Internet: Windows 95 Windows 98 Internet: Windows 98 Windows CE Internet: Windows CE Windows Family Internet: Windows Family Windows ME Internet: Windows ME Seite 1 von 138 eBook - Informations about Operating Systems Version: August 15, 2006 | Download: www.operating-system.org Windows NT 3.1 Internet: Windows NT 3.1 Windows NT 4.0 Internet: Windows NT 4.0 Windows Server 2003 Internet: Windows Server 2003 Windows Vista Internet: Windows Vista Windows XP Internet: Windows XP Apple - Company Internet: Apple - Company AT&T - Company Internet: AT&T - Company Be Inc. - Company Internet: Be Inc. - Company BSD Family Internet: BSD Family Cray Inc.
    [Show full text]
  • Introduzione Al Mondo Freebsd
    Introduzione al mondo FreeBSD Corso avanzato Netstudent Netstudent http://netstudent.polito.it E.Richiardone [email protected] maggio 2009 CC-by http://creativecommons.org/licenses/by/2.5/it/ The FreeBSD project - 1 ·EÁ un progetto software open in parte finanziato ·Lo scopo eÁ mantenere e sviluppare il sistema operativo FreeBSD ·Nasce su CDROM come FreeBSD 1.0 nel 1993 ·Deriva da un patchkit per 386BSD, eredita codice da UNIX versione Berkeley 1977 ·Per problemi legali subisce un rallentamento, release 2.0 nel 1995 con codice royalty-free ·Dalla release 5.0 (2003) assume la struttura che ha oggi ·Disponibile per x86 32 e 64bit, ia64, MIPS, ppc, sparc... ·La mascotte (Beastie) nasce nel 1984 The FreeBSD project - 2 ·Erede di 4.4BSD (eÁ la stessa gente...) ·Sistema stabile; sviluppo uniforme; codice molto chiaro, ordinato e ben commentato ·Documentazione ufficiale ben curata ·Licenza molto permissiva, spesso attrae aziende per progetti commerciali: ·saltuariamente esterni collaborano con implementazioni ex-novo (i.e. Intel, GEOM, atheros, NDISwrapper, ZFS) ·a volte no (i.e. Windows NT) ·Semplificazione di molte caratteristiche tradizionali UNIX Di cosa si tratta Il progetto FreeBSD include: ·Un sistema base ·Bootloader, kernel, moduli, librerie di base, comandi e utility di base, servizi tradizionali ·Sorgenti completi in /usr/src (~500MB) ·EÁ giaÁ abbastanza completo (i.e. ipfw, ppp, bind, ...) ·Un sistema di gestione per software aggiuntivo ·Ports e packages ·Documentazione, canali di assistenza, strumenti di sviluppo ·i.e. Handbook,
    [Show full text]
  • The Challenges of Dynamic Network Interfaces
    The Challenges of Dynamic Network Interfaces Brooks Davis The FreeBSD Project Seattle, WA brooks@{aero,FreeBSD}.org Abstract vices to the modern age of near complete dynamism. Following this history, the problems caused by this dynamism are discussed in detail. Then solutions to On early BSD systems, network interfaces were some of these problems are proposed and analyzed, static objects created at kernel compile time. To- and advice to implementers of userland applications day the situation has changed dramatically. PC is given. Finally, the issues are summarized and fu- Card, USB, and other removable buses allow hard- ture work is discussed. ware interfaces to arrive and depart at run time. Pseudo-device cloning also allows pseudo-devices to be created dynamically. Additionally, in FreeBSD and Dragonfly, interfaces can be renamed by the ad- 2 History ministrator. With these changes, interfaces are now dynamic objects which may appear, change, or dis- appear at any time. This dynamism invalidates a In early versions of UNIX, the exact set of devices number of assumptions that have been made in the on the system had to be compiled in to the kernel. If kernel, in external programs, and even in standards the administrator attempted to use a device which such as SNMP. This paper explores the history of was compiled in, but not installed, a panic or hang the transition of network interfaces from static to dy- was nearly certain. This system was easy to pro- namic. Issues raised by these changes are discussed gram and efficient to execute. Unfortunately, it was and possible solutions suggested.
    [Show full text]
  • PC-BSD 9 Turns a New Page
    CONTENTS Dear Readers, Here is the November issue. We are happy that we didn’t make you wait for it as long as for October one. Thanks to contributors and supporters we are back and ready to give you some usefull piece of knowledge. We hope you will Editor in Chief: Patrycja Przybyłowicz enjoy it as much as we did by creating the magazine. [email protected] The opening text will tell you What’s New in BSD world. It’s a review of PC-BSD 9 by Mark VonFange. Good reading, Contributing: especially for PC-BSD users. Next in section Get Started you Mark VonFange, Toby Richards, Kris Moore, Lars R. Noldan, will �nd a great piece for novice – A Beginner’s Guide To PF Rob Somerville, Erwin Kooi, Paul McMath, Bill Harris, Jeroen van Nieuwenhuizen by Toby Richards. In Developers Corner Kris Moore will teach you how to set up and maintain your own repository on a Proofreaders: FreeBSD system. It’s a must read for eager learners. Tristan Karstens, Barry Grumbine, Zander Hill, The How To section in this issue is for those who enjoy Christopher J. Umina experimenting. Speed Daemons by Lars R Noldan is a very good and practical text. By reading it you can learn Special Thanks: how to build a highly available web application server Denise Ebery with advanced networking mechanisms in FreeBSD. The Art Director: following article is the �nal one of our GIS series. The author Ireneusz Pogroszewski will explain how to successfully manage and commission a DTP: complex GIS project.
    [Show full text]
  • Absolute BSD—The Ultimate Guide to Freebsd Table of Contents Absolute BSD—The Ultimate Guide to Freebsd
    Absolute BSD—The Ultimate Guide to FreeBSD Table of Contents Absolute BSD—The Ultimate Guide to FreeBSD............................................................................1 Dedication..........................................................................................................................................3 Foreword............................................................................................................................................4 Introduction........................................................................................................................................5 What Is FreeBSD?...................................................................................................................5 How Did FreeBSD Get Here?..................................................................................................5 The BSD License: BSD Goes Public.......................................................................................6 The Birth of Modern FreeBSD.................................................................................................6 FreeBSD Development............................................................................................................7 Committers.........................................................................................................................7 Contributors........................................................................................................................8 Users..................................................................................................................................8
    [Show full text]
  • The Release Engineering of Freebsd 4.4
    The Release Engineering of FreeBSD 4.4 Murray Stokely [email protected] Wind River Systems Abstract different pace, and with the general assumption that they This paper describes the approach used by the FreeBSD re- have first gone into FreeBSD-CURRENT and have been lease engineering team to make production-quality releases thoroughly tested by our user community. of the FreeBSD operating system. It details the methodol- In the interim period between releases, nightly snap- ogy used for the release of FreeBSD 4.4 and describes the shots are built automatically by the FreeBSD Project build tools available for those interested in producing customized machines and made available for download from ftp: FreeBSD releases for corporate rollouts or commercial pro- //stable.FreeBSD.org. The widespread availabil- ductization. ity of binary release snapshots, and the tendency of our user community to keep up with -STABLE development with CVSup and “make world”[8] helps to keep FreeBSD- 1 Introduction STABLE in a very reliable condition even before the qual- ity assurance activities ramp up pending a major release. The development of FreeBSD is a very open process. Bug reports and feature requests are continuously sub- FreeBSD is comprised of contributions from thousands of mitted by users throughout the release cycle. Problem people around the world. The FreeBSD Project provides reports are entered into our GNATS[9] database through anonymous CVS[1] access to the general public so that email, the send-pr(1) application, or via a web-based form. others can have access to log messages, diffs between de- In addition to the multitude of different technical mailing velopment branches, and other productivity enhancements lists about FreeBSD, the FreeBSD quality-assurance mail- that formal source code management provides.
    [Show full text]
  • The Complete Freebsd
    The Complete FreeBSD® If you find errors in this book, please report them to Greg Lehey <grog@Free- BSD.org> for inclusion in the errata list. The Complete FreeBSD® Fourth Edition Tenth anniversary version, 24 February 2006 Greg Lehey The Complete FreeBSD® by Greg Lehey <[email protected]> Copyright © 1996, 1997, 1999, 2002, 2003, 2006 by Greg Lehey. This book is licensed under the Creative Commons “Attribution-NonCommercial-ShareAlike 2.5” license. The full text is located at http://creativecommons.org/licenses/by-nc-sa/2.5/legalcode. You are free: • to copy, distribute, display, and perform the work • to make derivative works under the following conditions: • Attribution. You must attribute the work in the manner specified by the author or licensor. • Noncommercial. You may not use this work for commercial purposes. This clause is modified from the original by the provision: You may use this book for commercial purposes if you pay me the sum of USD 20 per copy printed (whether sold or not). You must also agree to allow inspection of printing records and other material necessary to confirm the royalty sums. The purpose of this clause is to make it attractive to negotiate sensible royalties before printing. • Share Alike. If you alter, transform, or build upon this work, you may distribute the resulting work only under a license identical to this one. • For any reuse or distribution, you must make clear to others the license terms of this work. • Any of these conditions can be waived if you get permission from the copyright holder. Your fair use and other rights are in no way affected by the above.
    [Show full text]
  • UCL for Freebsd
    UCL for FreeBSD A universal config language for (almost) everything in FreeBSD Allan Jude -- ScaleEngine Inc. [email protected] twitter: @allanjude Introduction Allan Jude ● 13 Years as FreeBSD Server Admin ● FreeBSD docs committer (focus: ZFS, bhyve, ucl, xo) ● Co-Author of “FreeBSD Mastery: ZFS” ● Architect of the ScaleEngine CDN (HTTP and Video) ● Host of BSDNow.tv & TechSNAP.tv Podcasts ● Extensive work with Puppet to manage our 100+ servers in 35 data centers in 12 countries ● Lots of work with ZFS to manage large collections of videos, extremely large website caches, and the pkg mirror for PCBSD Impetus At EuroBSDCon 2014 in Sofia, Jordan Hubbard ([email protected]) gave the opening keynote Among the recommendations for how FreeBSD could be improved and made easier to manage was to move away from the ‘one config file format per utility’ Jordan, being from Apple, suggested the binary XML plists used by launchd. I really didn’t want the “one true format” to be XML. Why is this a “Good Idea”™? ● Includes allow for more readable, structured, and upgradable config files ● Overlays separate the defaults from your site wide settings, and from your local settings (like defaults/rc.conf rc.conf rc.conf.local) ● Includes allow packages to install default settings (installing apache or nginx adds its log rotation to newsyslog) ● ● What is UCL ● UCL -- universal configuration language ● Inspired by bind/nginx style configuration file that all sysadmins know well ● Fully compatible with JSON, but more liberal in what it accepts, so users do
    [Show full text]
  • Governance Structures of Free/Open Source Software Development
    Uitnodiging GOVERNANCE STRUCTURES OF FREE/OPEN 51 SOURCE SOFTWARE DEVELOPMENT Voor het bijwonen van de George Dafermos openbare verdediging van het Modularity theory makes a compelling argument: modular product design increases the potential number of persons that could work on a distributed project and has a positive effect proefschrift: on their labour productivity because it allows them to work independently of each other, with little or no need for central coordination. This doctoral dissertation sets out to put this argument to the test by studying a phenomenon that combines both scale and modularity: Governance Structures Free and open source software (FOSS) development. Its central question is: Does modularity mitigate the adverse effects of increasing scale in FOSS development? of Free/Open Source Software Development In exploring the effect of modularity and increasing scale on the dynamic of development of FreeBSD, a large and well-known FOSS project, over a period of fifteen years, the dissertation addresses several related empirical issues: How are FOSS projects organised? How are they governed? And most interestingly, how do they manage increasing scale? Does their ability to self-organise diminish as they grow larger, thereby necessitating hierarchical coordination? Op maandag 10 december 2012 om 15 uur precies in de Frans Source Software Development Source Software Governance Structures of Free/Open van Hasseltzaal van de Aula van de Technische Universiteit Delft, Mekelweg 5 te Delft The Next Generation Infrastructures Foundation George Dafermos represents an international consortium of knowledge institutions, market players and governmental bodies, which joined forces to cope with the challenges faced Governance Structures by today’s and tomorrow’s infrastructure systems.
    [Show full text]
  • Netmap: a Novel Framework for Fast Packet I/O
    netmap: a novel framework for fast packet I/O Luigi Rizzo,∗ Universita` di Pisa, Italy Abstract high rate raw packet I/O required by these applica- tions is not the intended target of general purpose OSes. Many applications (routers, traffic monitors, firewalls, Raw sockets, the Berkeley Packet Filter [14] (BPF), the etc.) need to send and receive packets at line rate even on AF SOCKET family, and equivalent APIs have been very fast links. In this paper we present netmap, a novel used to build all sorts of network monitors, traffic gen- framework that enables commodity operating systems erators, and generic routing systems. Performance, how- to handle the millions of packets per seconds traversing ever, is inadequate for the millions of packets per sec- 1..10 Gbit/s links, without requiring custom hardware or ond (pps) that can be present on 1..10 Gbit/s links. In changes to applications. search of better performance, some systems (see Sec- In building netmap, we identified and successfully re- tion 3) either run completely in the kernel, or bypass the duced or removed three main packet processing costs: device driver and the entire network stack by exposing per-packet dynamic memory allocations, removed by the NIC’s data structures to user space applications. Ef- preallocating resources; system call overheads, amor- ficient as they may be, many of these approaches depend tized over large batches; and memory copies, elimi- on specific hardware features, give unprotected access to nated by sharing buffers and metadata between kernel hardware, or are poorly integrated with the existing OS and userspace, while still protecting access to device reg- primitives.
    [Show full text]