infrastructure December 2006

infrastructure@lists.fedoraproject.org

34 participants
34 discussions

Introduction (and cfengine)
by David Douthitt 22 Dec '06

22 Dec '06

Hello! I am a UNIX System Administrator from the midwest (Wisconsin), and have worked with multiple open source and proprietary UNIX and Linux environments. (I'll skip over the UNIX and non-Red Hat experiences unless there is interest.) I've used Linux on non-Intel machines as well as Intel machines (PARISC and PowerPC specifically). I've used BSD systems as well, including on non-Intel architectures. I'm also certified: RHCE, SCSA, Linux+, and LPI Level 1. First time I used Linux was Slackware on a 386SX-16MHz machine with 4M of RAM; my oldest Red Hat is Red Hat 2 on a Linux User Group CD I still have around here. I also have 4.4BSD Lite and FreeBSD 2.2.6 and OpenBSD 2.6 on CDs..... I had my own floppy-based Linux distribution (Oxygen) based on the Linux Router Project. I have, in the past, also "built" my own Red Hat Enterprise systems without the benefit of an installer, and also in some cases from source RPMs alone (rather than binary RPMs). As far as cfengine goes, I contribute sporadically to the cfengine documentation, mostly proofreading and copyediting. I will be embarking on a book about cfengine shortly (after I complete the GNU screen book). I have run cfengine in a production environment and tend to be rather a voracious supporter of cfengine :-) I've not used puppet, but am a Ruby programmer and would like to learn more about it - it was designed after using cfengine and was designed to use the power of Ruby. I'm also a long-time Korn shell scripter - I used ksh for scripting before I used ksh as my shell! (I used csh then, but programmed in ksh.) I've also programmed in Perl. I've used Perl since the MSDOS 4.0 port, and Ruby since version 1.4. I also tend to be something of a programmer and programming languages nut; I've used FORTH (extensively and comprehensively), Smalltalk (dabble), LISP (moderate), and BASIC, C, Pascal, C++, COBOL, SNOBOL, all moderate to extensively. I contributed to the LCDd project briefly (mostly revamping the system to use GNU getopts). I'll be joining the documentation group at the same time. I'll also post a cfengine review shortly. -- David Douthitt HP-UX, Unixware, Linux, FreeBSD RHCE, SCSA, Linux+, LPIC-1 http://www.lulu.com/ssrat

3 4

Introduction / Xen Stuff
by Michael DeHaan 22 Dec '06

22 Dec '06

Hi, My name is Michael DeHaan, and I work in Red Hat's emerging technologies group in Raleigh, NC. I joined this list because (A) I really want to get more involved in the Fedora project beyond maintaining some extras packages, (B) I'm hoping that I can contribute some on the Xen and systems-management side of things, since I've heard infrastructure is going to start to do more with virtual machines. I also have a good deal of experience with other companies build systems (but not Red Hat's or Fedora's) and possibly I might be able to contribute some there also, or at least learn more about Fedora's setup. Mainly I'm interested in systems management and virtualization stuff though. Xen-wise, two things. I wrote and maintain a provisioning application called cobbler (http://cobbler.et.redhat.com), which potentially could be useful for provisioning Xen boxes in the infrastructure environment. Maybe. If not, maybe it can be extended, and feedback is awesome. If it's not useful, that's ok. The other thing is that I'm looking at ways to deploy Xen in easy ways for Fedora users, basically an open way of managing a /lot/ of Xen machines with integrated configuration management, central interfaces, and not a lot of setup or technical knowledge needed. That's all pretty new, though if some of that would be useful for infrastructure as it evolves, that would be pretty neat. Either way, I figure I can learn from ways that Fedora is using Xen just the same. Anyhow, hi. --Michael

2 1

Mailing List Troubles?
by David Douthitt 22 Dec '06

22 Dec '06

I sent my signup to the list using the *old-fashioned* way: I sent an email to fedora-infrastructure-list-request(a)redhat.com (and to fedora-documentation-list-request(a)redhat.com) with "subscribe" in the subject and mail with a standard signature (marked by dash-dash-space on a line). Both replied twice (for whatever reason!). I also seem to have sent off an HTML mail for my intro! Ack! I expected it to ask (at which point I'd use text only) but natch.... -- David Douthitt HP-UX, Unixware, Linux, FreeBSD RHCE, SCSA, Linux+, LPIC-1 http://www.lulu.com/ssrat

1 0

bcfg2
by Jeffrey Ollie 21 Dec '06

21 Dec '06

Another option to look into for configuration management: http://mricon.livejournal.com/360529.html Jeff

8 25

Introduction
by Ask Bjørn Hansen 21 Dec '06

21 Dec '06

Hi everyone! I've written another mail to the list I'm about to send, but I figured it'd be polite to introduce myself first. I think it's very nice how most contributors to this list have been doing that. I'm Ask Bjørn Hansen. My little consulting company develops neat application for companies large and small (lately mostly larger ones) and occasionally at conferences and to companies I give advice on how to architect web applications to scale[1]. More relevant for the work you guys are doing here, then I've been running many of the perl.org services for the Perl community since 1999 or so. We usually use RHEL (3 and 4), but lately a bit of Fedora too in addition to a FreeBSD system here and there. I also have a cabinet with Fedora servers for a startup I am, uh, starting. We might change those to RHEL5 at some point, but more likely then we'll get the Stateless Linux magic working instead and use that for managing all the systems (virtual and physical). I joined this list to absorb any tips and ideas for running a cluster of RH/Fedora boxes -- and ideally to offer my help, but of course I haven't had any time for that. - ask [1] http://develooper.com/talks/ -- http://develooper.com/ - http://askask.com/

2 1

Load test 1 - Sat Dec 23rd
by Mike McGrath 21 Dec '06

21 Dec '06

So the load test for the proxy servers is scheduled for this Saturday, Dec 23rd at 10:00 a.m. (GMT). It will involve at most, proxy1, proxy2, app1, and app2. Though will most likely not use both app servers. One of the proxy servers will be used as a client so none of the other RH systems (including the balancer for now) should be affected. Depending on the results of this test we will proceed with another test at a later date. The test should take no longer than 2 hours but may go 3. Stacy, if this doesn't work for you let me know, this test sounds pretty sane and shouldn't have any impact on anything other than fedora servers. It'll also happen during non-peak hours which is nice. -Mike

2 1

TurboGears and version control
by Mike McGrath 17 Dec '06

17 Dec '06

So many of us have started using TurboGears for Fedora infrastructure projects. My question is how should we be version controlling the source? Do we check in the entire project or just the related kid templates, configs, controller files, etc. When you create a project with one version of TurboGears does it work with other versions? How do we prevent mismatches? -Mike

2 1

gary_lerhaupt@dell.com Re: Undelivered Mail Returned to Sender
by Warren Togami 15 Dec '06

15 Dec '06

Matt, is this person no longer at Dell? Who do we reassign the package ownership to? Warren Togami wtogami(a)redhat.com Mail Delivery System wrote: > This is the Postfix program at host buildsys.fedoraproject.org. > > I'm sorry to have to inform you that your message could not > be delivered to one or more recipients. It's attached below. > > For further assistance, please send mail to <postmaster> > > If you do so, please include this problem report. You can > delete your own text from the attached returned message. > > The Postfix program > > <gary_lerhaupt(a)dell.com>: host smtp.ins.dell.com[143.166.83.183] said: 501 > #5.1.1 bad address gary_lerhaupt(a)dell.com (in reply to RCPT TO command) > > > ------------------------------------------------------------------------ > > Reporting-MTA: dns; buildsys.fedoraproject.org > X-Postfix-Queue-ID: AEB04152140 > X-Postfix-Sender: rfc822; buildsys(a)fedoraproject.org > Arrival-Date: Fri, 15 Dec 2006 16:32:59 -0500 (EST) > > Final-Recipient: rfc822; gary_lerhaupt(a)dell.com > Action: failed > Status: 5.0.0 > Diagnostic-Code: X-Postfix; host smtp.ins.dell.com[143.166.83.183] said: 501 > #5.1.1 bad address gary_lerhaupt(a)dell.com (in reply to RCPT TO command) > > > ------------------------------------------------------------------------ > > Subject: > Broken dependencies in Fedora Extras - 2006-12-15 > From: > Fedora Extras repoclosure <buildsys(a)fedoraproject.org> > Date: > Fri, 15 Dec 2006 21:32:59 -0000 > To: > gary_lerhaupt(a)dell.com, bugs.michael(a)gmx.net > > To: > gary_lerhaupt(a)dell.com, bugs.michael(a)gmx.net > > > This is an automated mail created by an experimental script. > Your following packages in the repository contain broken dependencies: > > package: dkms - 2.0.13-1.fc6.noarch from fedora-extras-development-ppc > unresolved deps: > kernel-devel >

2 1

IRC Meeting Log Posted - 12/14/2006
by Jeffrey Tadlock 15 Dec '06

15 Dec '06

The notes from yesterday's Infrastructure meeting on IRC have been posted to the Wiki: http://fedoraproject.org/wiki/Infrastructure/Meetings/2006-12-14 --Jeffrey

1 0

Web server torturing
by Ahmed Kamal 14 Dec '06

14 Dec '06

Hi, Due to the web server hit faced with FC6 release, some actions are being taken to minimize the chance of facing such issues again. One of the steps is to stress test our web-server infrastructure to measure our current and future capabilities. I'd like to run some tests on fp.o web-server, please let me know your thoughts/comments. Here are some details. Test Targets: 1- Measure our bare (no caching) maximum serving rate 2- Measure our cached serving rate, to assess the implemented caching efficiency 3- Gather numbers like (When do we get CPU saturated, RAM requirements ..). Possibly draw graphs (everyone thinks graphs are cool), the numbers should help us base future calculations on a solid basis 4- Future: Possibly implement a mechanism to cap the maximum connected clients to a specific server, to the maximum it can handle gracefully, to avoid killing a server Test Plan: 1- A script was written which uses apache's ab tool to stress the server. Script will run on the web-server host. 2- The script fires a total number of connections equal to ten times the maximum concurrency rate (to get good average, and avoid transients) 3- The concurrency rate is sweeped between 10 and 400 (my 1G-RAM machine swaps at about 100 connections)? any suggestions? 4- All ab output is recorded for future analysis 5- A monitoring thread is fired before ab is launched. The monitoring uses "top" to record load/cpu/ram/process information in log files as well 6- Tests are repeated with "ab -k" for enabling the HTTP keep-alive option. Not sure if this is needed, or if it will make much difference! comments? 7- Tests are done once with caching enabled and one more time without caching Please let me know your thoughts about the testing setup, should we be recording more data? should we be stressing the server in a different way, should we be testing some apache config options ... etc ? Thanks

10 23

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

infrastructure December 2006