infrastructure June 2011

infrastructure@lists.fedoraproject.org

37 participants
41 discussions

Patch for FAS site
by aruntheguy＠gmail.com 07 Jun '11

07 Jun '11

Hello Admins, I noticed that the buttons in various forms are aligned to left completely. It could be more UI friendly if the buttons are aligned just below the form text boxes. Hence I have created a small patch to move them under the textboxes, and maintain a uniform button size. The patch is attached with this mail. -- Regards P.Arunmozhi Twitter: @tecoholic Website: http://arunmozhi.in

1 1

secondary01 moved to new hardware
by Stephen John Smoogen 07 Jun '11

07 Jun '11

The move of the virtual machine secondary01.phx2.fedoraproject.org is completed. Data is now stored on the netapp and the OS is RHEL6. We will continue to move http/https access to the main download servers over time. -- Stephen J Smoogen. "The core skill of innovators is error recovery, not failure avoidance." Randy Nelson, President of Pixar University. "Let us be kind, one to another, for most of us are fighting a hard battle." -- Ian MacLaren

2 1

SSH changed to pubkey auth only
by Ricky Zhou 05 Jun '11

05 Jun '11

Hey, just a note to let everybody know, we just switched all of our machines to disallow SSH password authentication, so if you were previously using passwords to login anywhere, you'll need to use key auth now. Thanks, Ricky

5 6

Introduction
by Athmane Madjoudj 03 Jun '11

03 Jun '11

Hello everyone, I'm Linux Administrator / RHCE, I've good skills in php/python/bash and deployment and integration of various php-based applications: Drupal/Joomla/Wordpress/Moodle/phpBB and Linux-based services: Apache HTTPd, MySQL, Postfix/Exim/Dovecot, vsFTPd/pure-FTPd, ISC BIND Occasionally I do security auditing using various tools included in Fedora Security Lab, Metasploit framework and Python. As a first step I'm interesting to join 'fi-apprentice', I've already read [1] and some scripts and figs from 'fedora-infrastructure' git repo. [1] https://fedoraproject.org/wiki/Infrastructure/GettingStarted Contact info: Full name: Athmane Madjoudj Email: athmane AT fedoraproject.org IRC: athmane on Freenode (#fedora-qa, #fedora-admin, #fedora-noc, #epel) Fedora Account: athmane Timezone: UTC +01:00 Regards. -- Athmane Madjoudj

2 1

Some UI suggestions in the FAS site
by aruntheguy＠gmail.com 03 Jun '11

03 Jun '11

Hello Fedora Infra Team members, I just wanted to explore Fedora and visited the site. The motivation was enough to encourage me to pitch in and contribute whatever I can. So I went to register myself and that turned out to be both a discouraging as well as a encouraging experience. The captcha system was very bad and I had to spent about 45 mins before ricky (IRC Nick) did it for me. I found some major lapses in this particular thing. I was informed by ricky that things would be better in the next update, but let me give some suggestions that might help you in building the next system a better system. First regarding the captcha that did my day: 1. There wasn't ant documentation about the type of the captcha, whether it is a case sensitive or a case insensitive one. It could have been put straight away in the page or as a tool tip. 2. There wasn't a way to reload or get a alternate captcha image if one couldn't make out whats in there. 3. The captcha image itself was flawed. Generally its wise to put distortions in one color and the text in other color or at least different shade of the same color. 4. Finally there wasn't a audio captcha for people who get confused with distorted images. I accept this one as a luxury, but the previous three were very important. The user interface of the FAS site itself isn't that well built. To be exact, there could be more AJAXification of things. One doesn't need to reload the entire page for everything. A few suggestions here: 1. Almost all the buttons in all the pages are completely left aligned, which puts them farther left than the forms that contain the input fields themselves, its a good UX design if it could be put just under the fields, that is at the center of the page. 2. And the size of the buttons are completely dependent on the text it contains, again maintaining a common size for all buttons, except for really long text ones is a good practice, gives a good feel. 3. In the My account page, the first thing is Account Name, I think it is User Name. There might be some technical reason for the change there, but I feel it is a good thing to maintain a single naming convention throughout the site. 4. When browsing through the group list, the entire page is refreshed, it could be made to load dynamically through AJAX. 5. In a overall feel, the site seemed to be very rigid in UI, there could be some fluidity. I didn't go through the entire site, I might come up more suggestions as a browse through the pages, most of the above are very trivial things. I think this feed back would be helpful for the team to improve things. Thank you for the work done, and wish you for continual work. -- Regards P.Arunmozhi Twitter: @tecoholic Website: http://arunmozhi.in

3 2

OpenShift's Planet PaaS - Need Help and Pointers
by Jimmy Guerrero 03 Jun '11

03 Jun '11

Hello, Over here in OpenShift <http://openshift.redhat.com/app/> -land we are looking at standing up a PaaS-centric blog aggregation site called Planet PaaS, similar to: http://planet.fedoraproject.org/ http://planet.mysql.com/ http://planet.jboss.org/ We'd like to downplay the commercial stuff, and keep it focused on PaaS as it relates to open source and development frameworks, middleware and languages. We've been approached by some news/media outlets to co-brand a Planet PaaS site with us, but I'd prefer to have no overt commercial interests dominate the site. I'd also like to be impartial. Meaning yeah, VMWare and other competitors might turn up there, and that's ok as long as it is PaaS/OSS/Developer-centric. Anybody on this list interested in this project? I have some budget to make this happen. If you have any leads, please point me in the right direction. -- Jimmy 'PaaS Dude' Guerrero Sr Product Marketing Manager OpenShift | PaaS by Red Hat Twitter: @openshift

2 1

Proposal to fix Design Team group
by Paul W. Frields 02 Jun '11

02 Jun '11

Recently the Design team moved their materials to a git repo managed in Fedora Hosted. The old location served from /srv/groups on the fedorapeople.org box, however, likely has many incoming links from the wiki and elsewhere which will now be broken. Design team assets are used widely throughout the project. Getting things from git is a substantially high bar to set for a random contributor (more so with a random person looking to use the free assets we champion). Although there is a git browser, it's still much easier for someone to understand a simple web directory page, or even use existing instructions to hook their desktop to the team depot via SFTP. So I'd like to propose a way of making the original location work as a git clone with a timed pull that keeps the content fairly fresh. (0) (optional, and might already have been done) We might want to see if there is any packing that could be done on the Design team's git repo to minimize its size. (1) The quota for the designteam area should be raised to approximately 8 GB. Currently the full repo size is 4.8 GB and we should plan some room for growth. (2) I can set up the right puppet bits to do a timed pull of the repo into /srv/groups/designteam, although I may need some help if my access allows me to sudo on fedorapeople.org. (I have sysadmin{,-noc,-test,-insight} and designteam group membership.) I've done something similar for Insight now so I believe I can do the same without much trouble for this purpose. This should restore all broken links and also keep the content fresh for more casual consumers of this content. +/-1, and please counter-suggest if you're -1. Thanks for reading! -- Paul W. Frields http://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 http://redhat.com/ - - - - http://pfrields.fedorapeople.org/ Red Hat Summit/JBossWorld -- Register now! http://.theredhatsummit.com

3 3

Meeting *TOMORROW* 2011-06-02 at 1900 UTC in #fedora-meeting
by Derek Carter 01 Jun '11

01 Jun '11

The infrastructure team will be having it's weekly meeting tomorrow 2011-06-02 at 1900 UTC in #fedora-meeting on the freenode network. Suggested topics (suggested by whom): * New folks introductions * Upcoming Tasks/Items(nirik) * Outages Retrospectives * Mass upgrade improvement ideas. * Congrats Codeblock: http://blog.ianweller.org/2011/06/01/congrats-to-ricky-elrod/ * Meeting tagged tickets: https://fedorahosted.org/fedora-infrastructure/query?status=new&status=assi… Submit your agenda items, as tickets in the trac instance and send a note replying to this thread. More info here: https://fedoraproject.org/wiki/Infrastructure/Meetings#Meetings Thanks -- Derek aka goozbach

1 0

Mass reboots/updates improvements
by Kevin Fenzi 01 Jun '11

01 Jun '11

Greetings. I think we have some lessons learned and things we could improve based on the issues we ran into yesterday on the mass reboot/updates. ;) Issues/Observations: * We can't seem to complete everything in a 2 hour window. We should block out more time, and/or have things more organized. * The build system /mnt/koji issues are due to a guest that was moved from one machine to a new one, and then somehow started on both after reboots. ;( * There were a few cases of too many cooks doing things at once to a machine. * Some physical machines were poorly or not at all labeled in things like pdu's and serial consoles. * We need to be better about retiring machines. Sometimes it's hard to see what shouldn't be up or should be. Ideas/improvements: * I'd like to look at splitting all our hosts into 3 groups (based on who we need to notify about reboots or outages): a) End users will see/notice an outage if this machine is down/not working. b) Fedora package maintainers or contributors will notice if this machine/service is not working/down. c) Everything else. Including things that if they were single instances would fit in the above, but are spread out, so they can be rebooted/updated one at a time (ie, app servers, etc). I've made a tenative list with all our hosts in these groups: ~kevin/mass-reboot-list on puppet01. Please look and see if you see anything that looks wrong or needs adjusting. With this split out, we can do any machines in "c" as we like as long as we are careful, we can do 'b' machines if we announce to devel-announce and schedule a window and 'a' machines if we announce to the main fedora announce list and schedule a window. All the windows should be shorter than what we saw yesterday. * We might look at having a updates miester (czar?:) who would be the only one allowed to touch machines in a read/write way. By default everyone else is hands off unless the updates miester asks them to work on something. This would allow us to not interfere with each other or duplicate effort. * Seth is working on tooling to tell us anytime we have a virtual machine thats set to start on boot, but not started now, or not set to start on boot but started now. * We need to go and label things in all the pdu's etc. I can look at doing that and writing up a file somewhere with all the places a particular machine is. Then, the ones we can't find, we will fill in when smooge and I are out at phx2. * I have started a SOP for retiring machines. It needs a lot of work: https://fedoraproject.org/wiki/Infrastructure_retire_machine_SOP please modify and clean up. The goal should be making it very clear when a machine has been retired so we don't confuse it with anything active. There is a new rhel5 kernel out (yes, right after we applied the last one yesterday.), so I would suggest we look at implementing some or all of these that make sense for those updates. ;) Thoughts? Rants? more suggestions? kevin

1 0

retiring blogs.fedoraproject.org on 2011-07-01
by Kevin Fenzi 01 Jun '11

01 Jun '11

Greetings. There has been some confusion and misunderstanding around blogs.fedoraproject.org recently. This email will hopefully clear them up. At the Tempe Fudcon, the Fedora Infrastructure team determined that we should retire our blogs.fedoraproject.org service. The reasons for this were: * We currently have no infrastructure members who are driving support of this service. (Fixing bugs, following upstream or improving the service). * Wordpress has a long history of security issues, requiring frequent, custom patches or updates to keep it secure. * Many other sites out there are providing blog services as their core mission, likely resulting in a more feature-full and compelling offering. Some of these sites like wordpress.com are commited to free software, like Fedora is. * Usage of the service was quite low. We have 92 blogs, but only 39 of them had more than 5 posts. Only 6 of them had posts in the last month, and only 23 had posts in 2011. While we are open to interested parties stepping forward to help us maintain the service, currently there is no compelling reason to keep the service running. We will be happy to help any user wishing to transition their blog to another service. We will provide web site redirects and dumps of their posts. Please contact fedora infrastructure (infrastructure(a)lists.fedoraproject.org) with further questions or concerns. kevin

3 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

infrastructure June 2011