Re: [Server-devel] serious Raspbian WiFi flaw discovered: works for 10SEC then cycles OFF/ON

2018-01-01 Thread Peter Robinson
>> We need to report this to the Raspberry Pi Foundation to see if they can
>> fix it -- WiFi connections are repeatedly failing, sometimes even very soon
>> after booting, profoundly affecting Internet-in-a-Box 6.5 !
>>
>> But first a big Thanks In Advance to all who can reproduce this & offer
>> your own experiences/perspectives prior to Thursday's [*] call:
>>
>>http://minutes.iiab.io
>>
>> Here are 2 (known) ways to reproduce the WiFi bug with near certainty:
>>
>>https://github.com/iiab/iiab/issues/638#issuecomment-354639673
>
>
> CLARIF: I do *not* mean to blame Raspbian, as we seek a solution here!!
>
> At this this point it's entirely possible the underlying/root cause is
> dhcpcd...and/or even RPi3 firmware / WiFi module etc?

Is it reproducible with upstream kernel/firmware on something like an
up to date Fedora install, there was some new firmware pushed upstream
that fixed some CVEs in the Broadcom firmware (CVE-2016-0801,
CVE-2017-0561, CVE-2017-9417), I'm not sure which firmware's were
affected/updated but it might be worth checking to see if it's fixed
on other distros that are closer to upstream.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Apache 2.4.6 on CentOS and 2.4.10 on Debian/Raspbian

2017-05-26 Thread Peter Robinson
On Fri, May 26, 2017 at 1:37 AM, Adam Holt  wrote:
> Just FYI...
>
> Apache 2.4.6 was released July ~19, 2013 (used by IIAB/XSCE 6.2 on CentOS).
>
> Apache 2.4.10 was released Jule ~19, 2014 (used by IIAB/XSCE 6.2 on
> Debian/Raspbian).
>
> Apache 2.4.25 was release Dec ~19, 2016...if anybody knows any particularly
> important risks above that Internet-in-a-Box may face, please let us know!

This should be completely ignored in regards to, Red Hat in RHEL, and
as a direct result. CentOS manages the patches/CVEs in the version of
apache shipped and the risks are dealt with by the security team.
Details for all CVEs can be seen at
https://access.redhat.com/security/security-updates/
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] ssh.service error on CentOS 7.3

2017-05-26 Thread Peter Robinson
On Thu, May 25, 2017 at 9:07 PM, Adam Holt  wrote:
> IIAB/XSCE 6.2 is installed on CentOS 7.3 on a NUC, and this error message
> appears whenever I open a Terminal:
>
>Redirecting to /bin/systemctl status  ssh.service
>Unit ssh.service could not be found.
>
> Does anyone know if/where I can fix that to sshd.service instead of
> ssh.service?

It's sshd.service not ssh.server as the error indicates. Make sure you
have openssh-server package installed.

> PS Apache crashed for no obvious reason, in the middle of the night after
> about 24hrs, but this was recitified with "systemctl restart httpd.service"
> so I'll keep an eye on it.  Great to see that IIAB/XSCE 6.2 largely runs on
> CentOS (not just Debian) thanks to George Hunt's extremely hard work!
>
> ___
> Server-devel mailing list
> Server-devel@lists.laptop.org
> http://lists.laptop.org/listinfo/server-devel
>
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Fedora 24 released / Fedora 22 "end-of-life" July 19 2016

2016-09-22 Thread Peter Robinson
On Wed, Sep 7, 2016 at 5:26 PM, Adam Holt <h...@laptop.org> wrote:
> On Thu, Jun 23, 2016 at 6:23 AM, Peter Robinson <pbrobin...@gmail.com>
> wrote:
>>
>> On Thu, Jun 23, 2016 at 10:36 AM, Adam Holt <h...@laptop.org> wrote:
>> > Congrats to the F24 Team:
>> >
>> > The strategic question from a broad deployment perspective across the
>> > world's lower-middle class is increasingly Raspberry Pi support,
>> > according
>> > to so many grassroots/field groups I'm speaking with in 2016, very
>> > increasingly tempted to dump Fedora/CentOS for Raspbian, no matter what
>> > I
>> > tell them, so that learning technologies are not impeded by traditional
>> > education bureaucracy -- coming into schools thru the front/back and
>> > side
>> > doors.
>> >
>> > I personally hope this "only Raspbian can save us" sentiment is
>> > premature,
>> > in that I don't see the Raspbian ecosystem as being fully
>> > mature+resilient
>> > just yet -- and as such I *hope* CentOS (or Fedora, or Debian, or...)
>> > deliver increasingly competitive offerings on RPi 3, RPi 4, RPi 5 (or
>> > similar) into 2020~
>>
>> I have most of the bits in place for the RPi2/3 in F-24, just ran out
>> of time in the lead up to Beta to land the last bits. We will have
>> F-24 images for them soon and OOTB support in F-25.
>
>
> Thanks Peter!  Is this part of the F25 Alpha released last week and/or the
> F25 Beta expected in a month?

It's landed for Beta. Both the RPi2 and RPi3 both running 32-bit and a
single image for either them or any other device. More details will be
out in the coming days,

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Raspberry Pi 3

2016-09-22 Thread Peter Robinson
On Tue, Apr 5, 2016 at 3:49 PM, Tim Moody  wrote:
> For those who wish to experiment I have added several rpi3 images to
> http://xsce.org/downloads/xsce-release-6.0/rpi-images/

Fedora 25 will have Raspberry Pi support for both the RPi2 and RPi3 as
part of Beta due in a couple of weeks.

> These are based on the rpi2 images, so they are 32 bit, but they are
> configured to use the internal wifi as a hotspot.

In our builds (Fedora Minimal/Server etc editions) you'll be able to
dd the image out and boot it on either device without any changes.

> The Raspberry Pi folks also claim the 3 is 30% faster than the 2, even in 32
> bit mode, and have not yet announced plans for 64 bit support.

Having been testing both the 2 and 3 I can easily see this. It's
fairly basic... the v2 is 900mhz quad core, the v3 is 1.2ghz with 4
cores. So that's instantly 1/3 increase.

In RPi3 news I almost fell over when the firmware made it upstream
into linux-firmware the other day! Not all the drivers are upstream
yet though. I'm going to review how terrible and how many patches are
needed to try and get it landed for F-25 GA. Time will tell

http://git.kernel.org/cgit/linux/kernel/git/firmware/linux-firmware.git/commit/?id=c4c07a8d1128d50a5c2885ceea1abbebaa82f820
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Fedora 24 released / Fedora 22 "end-of-life" July 19 2016

2016-06-23 Thread Peter Robinson
On Thu, Jun 23, 2016 at 5:35 PM, George Hunt <georgejh...@gmail.com> wrote:
> Peter,
>
> What can you tell me about rpi3 bluetooth support in the F-24 images that
> will be coming out soon?

Haven't looked at it, we can't legally redistribute the firmware for
either the wifi or the BT. The initial release won't have OOTB support
for either (firmware or otherwise).

> By necessity, I've been playing with raspian, because there's been a lot of
> hype/hope for opportunistic device to device file sharing with rpi3.

Meh, it's never really taken off anywhere else

> On Thu, Jun 23, 2016 at 3:23 AM, Peter Robinson <pbrobin...@gmail.com>
> wrote:
>>
>> On Thu, Jun 23, 2016 at 10:36 AM, Adam Holt <h...@laptop.org> wrote:
>> > Congrats to the F24 Team:
>> >
>> > The strategic question from a broad deployment perspective across the
>> > world's lower-middle class is increasingly Raspberry Pi support,
>> > according
>> > to so many grassroots/field groups I'm speaking with in 2016, very
>> > increasingly tempted to dump Fedora/CentOS for Raspbian, no matter what
>> > I
>> > tell them, so that learning technologies are not impeded by traditional
>> > education bureaucracy -- coming into schools thru the front/back and
>> > side
>> > doors.
>> >
>> > I personally hope this "only Raspbian can save us" sentiment is
>> > premature,
>> > in that I don't see the Raspbian ecosystem as being fully
>> > mature+resilient
>> > just yet -- and as such I *hope* CentOS (or Fedora, or Debian, or...)
>> > deliver increasingly competitive offerings on RPi 3, RPi 4, RPi 5 (or
>> > similar) into 2020~
>>
>> I have most of the bits in place for the RPi2/3 in F-24, just ran out
>> of time in the lead up to Beta to land the last bits. We will have
>> F-24 images for them soon and OOTB support in F-25.
>>
>> P
>>
>> >
>> > From: Matthew Miller <mat...@fedoraproject.org>
>> > Date: Tue, Jun 21, 2016 at 9:47 AM
>> > Subject: Fedora 24 is here!
>> > To: annou...@lists.fedoraproject.org
>> >
>> > Today the Fedora Project is pleased to announce the general release of
>> > Fedora 24. Download it now from our Get Fedora site:
>> >
>> >  Workstation: https://getfedora.org/workstation/
>> >  Server:  https://getfedora.org/server/
>> >  Cloud:   https://getfedora.org/cloud/
>> >
>> >  Spins:   https://spins.fedoraproject.org/
>> >  Labs:https://labs.fedoraproject.org/
>> >  ARM: https://arm.fedoraproject.org/
>> >
>> >
>> > Another Step in the Fedora Journey
>> > --
>> >
>> > The Fedora Project has embarked on a great journey... redefining what
>> > an operating system should be for users and developers. Such innovation
>> > does not come overnight, and Fedora 24 is one big step on the road to
>> > the next generation of Linux distributions. But that does not mean that
>> > Fedora 24 is some "interim" release; there are great new features for
>> > Fedora users to deploy in their production environments right now!
>> >
>> >
>> > Workstation
>> > ---
>> >
>> > The Fedora 24 Workstation release features GNOME 3.20, with many
>> > usability improvements such as easier input device and printer
>> > settings, a better search interface, shortcut windows for keyboard
>> > commands, and more convenient music controls.
>> >
>> > Flatpak (formerly xdg-app) is another building-block feature, with
>> > Software able to track installed Flatpaks and adding more features in
>> > the future as the technology develops. The Software app has also grown
>> > features to provide a full system upgrade directly from the desktop
>> > from one Fedora release to the next, and the ability to provide
>> > labeling as well as reviews of available software.
>> >
>> > Fedora 24 continues our work on the X replacement, Wayland, a
>> > next-generation graphics stack. Although this release will not default
>> > to Wayland, it includes many improvements and is available as an option
>> > for users to try out, and potentially will be the default stack in
>> > Fedora 25.
>> >
>> >
>> > Server
>> > --
>> >
>> > Fedora 24 Server edition is more streamlined and introduces more
>> > modularity, which will become a major fa

Re: [Server-devel] Fedora 24 released / Fedora 22 "end-of-life" July 19 2016

2016-06-23 Thread Peter Robinson
On Thu, Jun 23, 2016 at 10:36 AM, Adam Holt  wrote:
> Congrats to the F24 Team:
>
> The strategic question from a broad deployment perspective across the
> world's lower-middle class is increasingly Raspberry Pi support, according
> to so many grassroots/field groups I'm speaking with in 2016, very
> increasingly tempted to dump Fedora/CentOS for Raspbian, no matter what I
> tell them, so that learning technologies are not impeded by traditional
> education bureaucracy -- coming into schools thru the front/back and side
> doors.
>
> I personally hope this "only Raspbian can save us" sentiment is premature,
> in that I don't see the Raspbian ecosystem as being fully mature+resilient
> just yet -- and as such I *hope* CentOS (or Fedora, or Debian, or...)
> deliver increasingly competitive offerings on RPi 3, RPi 4, RPi 5 (or
> similar) into 2020~

I have most of the bits in place for the RPi2/3 in F-24, just ran out
of time in the lead up to Beta to land the last bits. We will have
F-24 images for them soon and OOTB support in F-25.

P

>
> From: Matthew Miller 
> Date: Tue, Jun 21, 2016 at 9:47 AM
> Subject: Fedora 24 is here!
> To: annou...@lists.fedoraproject.org
>
> Today the Fedora Project is pleased to announce the general release of
> Fedora 24. Download it now from our Get Fedora site:
>
>  Workstation: https://getfedora.org/workstation/
>  Server:  https://getfedora.org/server/
>  Cloud:   https://getfedora.org/cloud/
>
>  Spins:   https://spins.fedoraproject.org/
>  Labs:https://labs.fedoraproject.org/
>  ARM: https://arm.fedoraproject.org/
>
>
> Another Step in the Fedora Journey
> --
>
> The Fedora Project has embarked on a great journey... redefining what
> an operating system should be for users and developers. Such innovation
> does not come overnight, and Fedora 24 is one big step on the road to
> the next generation of Linux distributions. But that does not mean that
> Fedora 24 is some "interim" release; there are great new features for
> Fedora users to deploy in their production environments right now!
>
>
> Workstation
> ---
>
> The Fedora 24 Workstation release features GNOME 3.20, with many
> usability improvements such as easier input device and printer
> settings, a better search interface, shortcut windows for keyboard
> commands, and more convenient music controls.
>
> Flatpak (formerly xdg-app) is another building-block feature, with
> Software able to track installed Flatpaks and adding more features in
> the future as the technology develops. The Software app has also grown
> features to provide a full system upgrade directly from the desktop
> from one Fedora release to the next, and the ability to provide
> labeling as well as reviews of available software.
>
> Fedora 24 continues our work on the X replacement, Wayland, a
> next-generation graphics stack. Although this release will not default
> to Wayland, it includes many improvements and is available as an option
> for users to try out, and potentially will be the default stack in
> Fedora 25.
>
>
> Server
> --
>
> Fedora 24 Server edition is more streamlined and introduces more
> modularity, which will become a major factor in future Fedora releases,
> even as unnecessary packages were removed and the installer has a
> smaller footprint.
>
> FreeIPA 4.3 is a major feature for Fedora 24 Server.
> FreeIPA is an integrated security information management solution. This
> new version of FreeIPA features simplified replica installation and
> improved replication technology management.
>
>
> Cloud
> -
>
> Fedora is on its way to being the best platform for containerized
> applications, from base Fedora container images to a full-featured
> platform as a service to run and manage them.
>
> As we continue on this part of the journey, we are packaging OpenShift
> Origin so it is easy to deploy. OpenShift Origin is a Platform as a
> Service system based around Kubernetes, a production-grade container
> orchestration project. OpenShift Origin is optimized for application
> development and deployment. Origin makes it easy for developers to get
> started building applications in containers and for operators to manage
> them.
>
> While not shipped in Fedora 24, per se, we have new infrastructure for
> developing container images with applications layered on top of the
> base Fedora Docker image. Fedora Developers will also see a layered
> image build service, which provides tools for Fedora contributors to
> start creating and shipping layered container images in Fedora 25 and
> beyond.
>
>
> Spins and Labs
> --
>
> Fedora Spins and Labs are alternative Fedora versions that offer
> additional desktop environments, or other custom collections of
> software, alongside the three editions that are the primary focus for
> the project.
>
> Our Spins make it easy for people to use other desktop environments.
> 

Re: [Server-devel] Fwd: Announcing the release of Fedora 24 Beta for aarch64!

2016-05-10 Thread Peter Robinson
On Tue, May 10, 2016 at 9:39 PM, Adam Holt <h...@laptop.org> wrote:
> On May 10, 2016 3:21 PM, "Peter Robinson" <pbrobin...@gmail.com> wrote:
>>
>> We do have initial pine64 support, I plan to improve it more for GA and
>> have a disk image for use.
>
> Great News!
>
> Tangentially: are there opinions emerging on the physical ruggedizability of
> 2GB Pine64 v. 1GB RPi3 (incl WiFi) for developing world education, and harsh
> environmental/ergonomic conditions, as more people start banging on both?

I'm sure I've said it before, it's not high quality. My alpha board
arrived bent (still works somehow) but the hackaday review [1] sums up
my opinion quite well. This isn't a robust device, and with up to 2Gb
of RAM there's a lot more ARMv7 devices that offer better value for
money. At a max of 2Gb of RAM a 64 bit chip offers absolutely ZERO
advantage and it's built for a price and I'm sorry but "harsh
environment" is not anywhere in that. Honestly a BeagleBone would do
better and has much better build and likely better performance for not
much more price (yes, I do class double the price in this class as not
much more)

[1] http://hackaday.com/2016/04/21/pine64-the-un-review/

>> Peter
>>
>> On 10 May 2016 18:44, "Adam Holt" <h...@laptop.org> wrote:
>>>
>>> Not sure the 64-bit implications of this for RPi3 and Pine64, but FYI.
>>>
>>> -- Forwarded message --
>>> From: "Peter Robinson" <pbrobin...@gmail.com>
>>> Date: May 10, 2016 1:34 PM
>>> Subject: Announcing the release of Fedora 24 Beta for aarch64!
>>> To: <a...@lists.fedoraproject.org>,
>>> <devel-annou...@lists.fedoraproject.org>,
>>> <annou...@lists.fedoraproject.org>, <second...@lists.fedoraproject.org>
>>> Cc:
>>>
>>>> The Fedora 24 Beta for aarch64 is here, on schedule for our planned June
>>>> final
>>>> release. For Beta we have added Cloud and Docker base images.
>>>> Download the prerelease from our Get Fedora site:
>>>>
>>>> -   Get Fedora 24 Beta Server: make use of the very latest server-based
>>>> technologies available in the open source community
>>>>
>>>> https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/Server/aarch64/
>>>>
>>>> -   Get Fedora 24 Beta Cloud: build scale-out computing and utilize the
>>>> next
>>>> generation of container deployment technology
>>>>
>>>> https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/CloudImages/aarch64/
>>>>
>>>> https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/Docker/aarch64/images/
>>>>
>>>> What is the Beta release?
>>>> -
>>>>
>>>> The Beta release contains all the exciting features of Fedora 24's
>>>> editions in a form that anyone can help test. This testing, guided by
>>>> the Fedora QA team, helps us target and identify bugs from the Alpha
>>>> version. When most of these bugs are fixed, we make a Beta release
>>>> available. A Beta release is code-complete and bears a very strong
>>>> resemblance to the third and final release. The final release of Fedora
>>>> 24 is expected in June. We need your help to make Fedora 24 the best
>>>> yet. Please take some time to download and try out the Beta and make
>>>> sure the things that are important to you are working. If you find a
>>>> bug, please report it – every bug you uncover is a chance to improve the
>>>> experience for millions of Fedora users worldwide. This is a great
>>>> opportunity for non-programmers to contribute back to fedora. Together,
>>>> we can make Fedora rock-solid. We have a culture of adding new features
>>>> to software and pushing fixes to the upstream developers at the same
>>>> time. This means your feedback will help improve not only Fedora but
>>>> Linux and free software on the whole.
>>>>
>>>> -   <https://fedoraproject.org/wiki/Releases/24/Schedule>
>>>> -   <https://fedoraproject.org/wiki/How_to_file_a_bug_report>
>>>>
>>>> Fedora-Wide Changes
>>>> ---
>>>>
>>>> Under the hood, glibc has moved to 2.23. This update includes better
>>>> performance, increased security, bugfixes, improvements to POSIX
>>>> compliance, and additional locales. The new library is backwards
>>

Re: [Server-devel] Fwd: Announcing the release of Fedora 24 Beta for aarch64!

2016-05-10 Thread Peter Robinson
We do have initial pine64 support, I plan to improve it more for GA and
have a disk image for use.

Peter
On 10 May 2016 18:44, "Adam Holt" <h...@laptop.org> wrote:

> Not sure the 64-bit implications of this for RPi3 and Pine64, but FYI.
> -- Forwarded message ------
> From: "Peter Robinson" <pbrobin...@gmail.com>
> Date: May 10, 2016 1:34 PM
> Subject: Announcing the release of Fedora 24 Beta for aarch64!
> To: <a...@lists.fedoraproject.org>, <devel-annou...@lists.fedoraproject.org>,
> <annou...@lists.fedoraproject.org>, <second...@lists.fedoraproject.org>
> Cc:
>
> The Fedora 24 Beta for aarch64 is here, on schedule for our planned June
> final
> release. For Beta we have added Cloud and Docker base images.
> Download the prerelease from our Get Fedora site:
>
> -   Get Fedora 24 Beta Server: make use of the very latest server-based
> technologies available in the open source community
>
> https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/Server/aarch64/
>
> -   Get Fedora 24 Beta Cloud: build scale-out computing and utilize the
> next
> generation of container deployment technology
>
> https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/CloudImages/aarch64/
>
> https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/Docker/aarch64/images/
>
> What is the Beta release?
> -
>
> The Beta release contains all the exciting features of Fedora 24's
> editions in a form that anyone can help test. This testing, guided by
> the Fedora QA team, helps us target and identify bugs from the Alpha
> version. When most of these bugs are fixed, we make a Beta release
> available. A Beta release is code-complete and bears a very strong
> resemblance to the third and final release. The final release of Fedora
> 24 is expected in June. We need your help to make Fedora 24 the best
> yet. Please take some time to download and try out the Beta and make
> sure the things that are important to you are working. If you find a
> bug, please report it – every bug you uncover is a chance to improve the
> experience for millions of Fedora users worldwide. This is a great
> opportunity for non-programmers to contribute back to fedora. Together,
> we can make Fedora rock-solid. We have a culture of adding new features
> to software and pushing fixes to the upstream developers at the same
> time. This means your feedback will help improve not only Fedora but
> Linux and free software on the whole.
>
> -   <https://fedoraproject.org/wiki/Releases/24/Schedule>
> -   <https://fedoraproject.org/wiki/How_to_file_a_bug_report>
>
> Fedora-Wide Changes
> ---
>
> Under the hood, glibc has moved to 2.23. This update includes better
> performance, increased security, bugfixes, improvements to POSIX
> compliance, and additional locales. The new library is backwards
> compatible with the version of glibc that was shipped in Fedora 23.
> We've also updated the system compiler to GCC 6 and rebuilt all of our
> packages with it, providing greater code optimization and improved
> program error catching.
>
> Server
> --
>
> Fedora 24 beta server edition has also been more streamlined. Unnecessary
> packages were removed and the installer has a smaller footprint. FreeIPA
> 4.3, an integrated security information management solution is now
> included. The installation of replicas is streamlined by adding a
> replica promotion method for new installs. A new topology plugin has
> also been added to this version of FreeIPA that automatically manages
> new replication segment creation. An effective replica topology
> visualization tool is now available in the webUI.
>
> Cloud
> -
>
> We are working hard to make Fedora the best platform for containerized
> applications, from base Fedora container images to a full-featured
> platform as a service to run and manage them. To meet this goal, we are
> packaging OpenShift Origin so it is easy to deploy. OpenShift Origin is
> a distribution of Kubernetes, a container cluster manager from Google.
> It is optimized for enterprise application development and deployment.
> Origin makes it easy for developers to get started building applications
> in containers and for operators to manage them.
>
> Issues and Details
> --
>
> This is a Beta release. As such, we expect that you may encounter bugs
> or missing features. To report issues encountered during testing,
> contact the Fedora QA team via the mailing list or in #fedora-qa on
> Freenode. As testing progresses, common issues are tracked on the Common
> F24 Bugs page.
>
> -   <h

Re: [Server-devel] [XSCE] Re: [UKids] Raspberri Pi/clone(s) most ruggedizable for OLPC fieldwork?

2016-04-06 Thread Peter Robinson
Not sure how they do that. If it's done with something like ansible it
would be very easy to migrate to new releases.

Peter

On Wed, Apr 6, 2016 at 2:23 PM, Tim Moody <t...@timmoody.com> wrote:
> Jerry and George spent a lot of time migrating legacy XS code from F18 to 
> F22.  There is more work to get to F24.  I agree that we want to get there.
>
>> -Original Message-
>> From: Peter Robinson [mailto:pbrobin...@gmail.com]
>> Sent: Wednesday, April 06, 2016 6:42 AM
>> To: Tim Moody <t...@timmoody.com>
>> Cc: xsce-de...@googlegroups.com; Adam Holt <h...@laptop.org>; server-
>> devel <server-devel@lists.laptop.org>
>> Subject: Re: [Server-devel] [XSCE] Re: [UKids] Raspberri Pi/clone(s) most
>> ruggedizable for OLPC fieldwork?
>>
>> On Sat, Apr 2, 2016 at 8:15 PM, Tim Moody <t...@timmoody.com> wrote:
>> > For xsce I'd start with f22.  You can look at the rpi images on
>> > xsce.org/downloads
>>
>> They're 32 bit ARMv7 are they not? Personally I'd be starting with Fedora 24
>> as you'll have support until July 2017, instead of the 3 or so months left 
>> for for
>> F-22. Also for an aarch64 device F-23+ is highly recommended.
>>
>> > Sent from my Samsung Galaxy smartphone.
>> >
>> >
>> >  Original message 
>> > From: Alex Perez <ape...@alexperez.com>
>> > Date: 4/2/2016 2:12 PM (GMT-05:00)
>> > To: Adam Holt <h...@laptop.org>
>> > Cc: server-devel <server-devel@lists.laptop.org>,
>> > xsce-de...@googlegroups.com
>> > Subject: [XSCE] Re: [UKids] Raspberri Pi/clone(s) most ruggedizable
>> > for OLPC fieldwork?
>> >
>> >
>> > On Feb 6, 2016, at 1:10 PM, Adam Holt <h...@laptop.org> wrote:
>> >
>> > Will the $15 http://pine64.com (which just raised $1.7M) become
>> > genuinely productizable for truly hassle-free field use with 128GB
>> > MicroSD cards by late 2016?  On the bright side, it accommodates 128GB
>> > maximum, which is exactly what we need on the high end in 2016.
>> >
>> >
>> > Looping back on this...
>> >
>> > My PINE64 arrived in the mail today. Mine is the 2GB RAM variant, with
>> > gigabit ethernet, but no bluetooth or Wi-Fi (which is an add-in
>> > module)
>> >
>> >  I actually have a couple of 200GB microSD cards that I just acquired,
>> > and I’ll be putting it through its paces. The SanDisk
>> > SDSDQUAN-200G-G4A retails for $249, but the actual street prices are
>> > $80 (on Amazon)
>> >
>> > I suspect it will work fine in the PINE64. Does anyone here have any
>> > recommendations for putting XSCE through its paces, or otherwise
>> > stress-testing the install, besides normal benchmarking of static HTTP
>> > page load performance.
>> >
>> > Regards,
>> > Alex Perez
>> >
>> >
>> > ___
>> > Server-devel mailing list
>> > Server-devel@lists.laptop.org
>> > http://lists.laptop.org/listinfo/server-devel
>> >
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] [XSCE] Re: [UKids] Raspberri Pi/clone(s) most ruggedizable for OLPC fieldwork?

2016-04-06 Thread Peter Robinson
On Sat, Apr 2, 2016 at 8:15 PM, Tim Moody  wrote:
> For xsce I'd start with f22.  You can look at the rpi images on
> xsce.org/downloads

They're 32 bit ARMv7 are they not? Personally I'd be starting with
Fedora 24 as you'll have support until July 2017, instead of the 3 or
so months left for for F-22. Also for an aarch64 device F-23+ is
highly recommended.

> Sent from my Samsung Galaxy smartphone.
>
>
>  Original message 
> From: Alex Perez 
> Date: 4/2/2016 2:12 PM (GMT-05:00)
> To: Adam Holt 
> Cc: server-devel ,
> xsce-de...@googlegroups.com
> Subject: [XSCE] Re: [UKids] Raspberri Pi/clone(s) most ruggedizable for OLPC
> fieldwork?
>
>
> On Feb 6, 2016, at 1:10 PM, Adam Holt  wrote:
>
> Will the $15 http://pine64.com (which just raised $1.7M) become genuinely
> productizable for truly hassle-free field use with 128GB MicroSD cards by
> late 2016?  On the bright side, it accommodates 128GB maximum, which is
> exactly what we need on the high end in 2016.
>
>
> Looping back on this...
>
> My PINE64 arrived in the mail today. Mine is the 2GB RAM variant, with
> gigabit ethernet, but no bluetooth or Wi-Fi (which is an add-in module)
>
>  I actually have a couple of 200GB microSD cards that I just acquired, and
> I’ll be putting it through its paces. The SanDisk SDSDQUAN-200G-G4A retails
> for $249, but the actual street prices are $80 (on Amazon)
>
> I suspect it will work fine in the PINE64. Does anyone here have any
> recommendations for putting XSCE through its paces, or otherwise
> stress-testing the install, besides normal benchmarking of static HTTP page
> load performance.
>
> Regards,
> Alex Perez
>
>
> ___
> Server-devel mailing list
> Server-devel@lists.laptop.org
> http://lists.laptop.org/listinfo/server-devel
>
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] [UKids] Raspberri Pi/clone(s) most ruggedizable for OLPC fieldwork?

2016-04-06 Thread Peter Robinson
On Sat, Apr 2, 2016 at 7:23 PM, Adam Holt  wrote:
> Awesome we can start banging on this long-awaited HW!.
>
> Who can recommend the best/emerging/viable Fedora 22 vs. 23 vs. 24 options
> to Alex?

Fedora 24 definitely here, I'll actually be producing aarch64 images
shortly for this which will include a level of support for the PINE64,
we don't get have USB or ethernet sadly, but I hope to have at least
usb soon.

I would note that, at least for the pre prod model I have, this device
is by no means "ruggedized"  the board even has a bend in it from
postage.

Peter

> On Sat, Apr 2, 2016 at 2:12 PM, Alex Perez  wrote:
>>
>>
>> On Feb 6, 2016, at 1:10 PM, Adam Holt  wrote:
>>
>> Will the $15 http://pine64.com (which just raised $1.7M) become genuinely
>> productizable for truly hassle-free field use with 128GB MicroSD cards by
>> late 2016?  On the bright side, it accommodates 128GB maximum, which is
>> exactly what we need on the high end in 2016.
>>
>>
>> Looping back on this...
>>
>> My PINE64 arrived in the mail today. Mine is the 2GB RAM variant, with
>> gigabit ethernet, but no bluetooth or Wi-Fi (which is an add-in module)
>>
>>  I actually have a couple of 200GB microSD cards that I just acquired, and
>> I’ll be putting it through its paces. The SanDisk SDSDQUAN-200G-G4A retails
>> for $249, but the actual street prices are $80 (on Amazon)
>>
>> I suspect it will work fine in the PINE64. Does anyone here have any
>> recommendations for putting XSCE through its paces, or otherwise
>> stress-testing the install, besides normal benchmarking of static HTTP page
>> load performance.
>>
>> Regards,
>> Alex Perez
>>
>> --
>> Unsung Heroes of OLPC, interviewed live @ http://unleashkids.org !
>
>
> ___
> Server-devel mailing list
> Server-devel@lists.laptop.org
> http://lists.laptop.org/listinfo/server-devel
>
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Rapberri Pi/clone(s) most ruggidizable for OLPC fieldwork?

2016-02-07 Thread Peter Robinson
> You are the expert here (on the low end especially, with $5 Raspberri Pi
> Zeros!) leading OLE refugee camp deployments across many countries.  Whereas
> most OLPC-like schools I talk to want to spend $100 (or more) for a
> mini-server that's truly resilient for years in tropical environments,
> against rodents/humidity, untrained-operators especially, and oh yeah kids
> smacking them around, stealing SD cards etc ;)
>
> (Yes some of us proudly confess we use XO-1.5 and similar mini-servers on
> the low end, e.g. in many after-school programs, despite Tony's friendly
> fundamentalism that XO's should not be allowed to serve others nearby,
> recycling these XO-1.5's into family/mentoring learning environments is a
> huge win, but more about that another day :-)
>
> Looking into the future, so many of these community deployment leaders want
> something beefier right now.  Sometimes even a hard disk drive "for just $50
> more" (that's the fantasy anyway!)  Even now that 4 or 8 GB RAM appears
> essentially free--it does not fit on most of these tiny motherboards.  Nor
> does a 1TB hard disk drive fit, that many of these groups dream of (and are
> generally happy to shove in a live/spare HDD inside, when the prior one
> dies!)
>
> So, in conclusion we're exactly 4 years into RPi era (began February 2012)
> and showing a lot of maturity^h^hgrayhair now :)  I assume there are a
> gazillion Raspberry Pi clones -- the central question is which ones we
> should narrow down upon, now that push has come to shove in 2016, serving
> educational microserver applications (school and non-schools) across those
> ~4 billion "offline" people now asking us for that:
>
> Will the $15 http://pine64.com (which just raised $1.7M) become genuinely
> productizable for truly hassle-free field use with 128GB MicroSD cards by
> late 2016?  On the bright side, it accommodates 128GB maximum, which is
> exactly what we need on the high end in 2016.
> Or are traditional RPi 2 clones, such as Banana Pi's and this $40 Korean
> unit more practical in the short-term?
> http://www.computerworld.com/article/3030251/computer-hardware/make-a-40-linux-or-android-pc-with-new-raspberry-pi-2-rival.html
> Understanding full well this is as much a plastics / industrial design
> question / economies-of-scale question first-and-foremost, more than a
> computer specs question per se.  Still today, few people appreciate (or even
> understand) how much OLPC's contribution a decade ago was a revolution in
> plastics/maintainability/droppability.  When Apple/Google are daily throwing
> gigahertz specs at us, that marketing-centric misunderstanding is of course
> only natural :/
>
> Etc, Thanks in advance to those who've done so much more RPi field research
> and can share!!

So having read all of the above I'm not sure what you're asking. Do
you want a device that does end user sugar style interface or do you
want a device for a server?

If the later I suggest two things that should be looked for when
looking for a cheap ARM device to do server:
1) SoC attached network (100Mb or Gb)
2) SoC attached SATA port

Without those two you generally get terrible performance. The
Raspberry Pi is terrible for this as everything is USB attached
through a single (buggy!) USB controller. In the case of things like
the PINE64 above it has a SoC attached network but not storage.

As is stands at the moment some of the best cheap devices for server
style devices is AllWinner A20 devices (CubieTruck, BananaPi and
friends) and i.MX6 devices (Wandboard, CuBox-i and friends)

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Raspberry Pi/clone(s) most ruggedizable for OLPC fieldwork?

2016-02-07 Thread Peter Robinson
On Sun, Feb 7, 2016 at 2:07 PM, Adam Holt <h...@laptop.org> wrote:
> On Feb 7, 2016 3:22 AM, "Peter Robinson" <pbrobin...@gmail.com> wrote:
>> things like
>> the PINE64 above it has a SoC attached network but not storage.
>
> Both SATA (real TB+ disks) and Ethernet (external Wi-Fi AP antennae) are
> icing-on-the-cake we will both strongly consider.
>
>> As is stands at the moment some of the best cheap devices for server
>> style devices is AllWinner A20 devices (CubieTruck, BananaPi and
>> friends) and i.MX6 devices (Wandboard, CuBox-i and friends)
>
> Hugely helpful.
>
> Key criterion for offline/remote deployments: does this accept 128GB MicroSD
> cards, so 2016's developing world $50-100 "knowledge hotspots" increasingly
> now become very real?  (Aside: 256GB MicroSD cards will be part of this well
> before 2020, apparently beyond the capability of most of these SoC's.)
>
> Peter, does Fedora 24 have a shot to one day run on the "$19" Pine64 Plus?!
> Even if it's ambiguous whether it can truly contain 2GB RAM as advertised,
> Pine64 claims to run up to 70C which is very promising if true.  ($15 Pine64
> contains 512MB, and $19 "Pine64 Plus" contains 1GB RAM.  Their 2GB RAM story
> is very attractive, but may be marketing vaporware for now?)

Yes, I've got one awaiting for me on my return to London. Kernel isn't
upstream, nor is u-boot, I'm not sure how big the patches are, I'm
hoping it'll all be landable in F-24.

> Or...would you recommend other ruggedized platforms to run Fedora, for
> schools/libraries/clinics needing this in place by January 1st 2017?
> (Thankfully size does not matter.  Cubox is very cute, and we will use it if
> it's the most rugged, but physically larger units are also fine too.
> Certainly Fedora remains a priority for now, given schoolserver.org's
> obvious OLPC legacy +)

Basically I'd want a specs set, I'm not sure why you'd want to use a
128Gb SD card over an actual SSD or HDD, the later are a lot more
robust. There's literally 100s of possible devices that would possibly
meet your needs, what would be great is a list of must haves and a
list of nice to haves and from that I could give a list of possible
options.
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Two (minimal) goals for a kernel for XO1.5 based upon fc22?

2015-05-28 Thread Peter Robinson
On Thu, May 28, 2015 at 4:17 AM, Samuel Greenfeld sam...@greenfeld.org wrote:
 Given I looked into this before with F22 Beta(*), I ran OLPC OS Builder
 tonight, excluding nothing from all repositories, and let it loose.

Funny! I was doing exactly the same yesterday for the 1.5!

 There were a few more minor issues found with the build process.  But I
 managed to build an image with the stock Fedora 4.x kernel, and primarily
 F22 parts.

Can we create a f22 branch on the o-o-b git repo so we can all
collaborate on this from one central place?

 However the stock Fedora kernel does not have a hardcoded kernel command
 line like OLPC kernels do, and trying to manually provide the command line
 to boot a Fedora kernel still hung for me after OFW stated the ramdisk was
 being loaded.

OK, got the command line or overview of what's required here? I have a
branch of a multi platform MMP2 (1.75) kernel I was trying to get to
boot on the 1.75 and having similar issues.

 I may look into this a bit more before resorting back to OLPC's 3.10 kernel
 and seeing what works with that.

 (*) http://lists.sugarlabs.org/archive/sugar-devel/2015-April/050035.html

 On Wed, May 27, 2015 at 6:28 PM, James Cameron qu...@laptop.org wrote:

 On Wed, May 27, 2015 at 06:31:28AM -0700, George Hunt wrote:
  If I set just the goals of getting keyboard input, and display output,
  what
  problems will I face trying to use defconfig_xo1.5 from
  [1]dev.laptop.org
  (x86-3.3 branch)?
 
  Does anyone have a config file that works on the XO1.5 on a Fedora
  release
  later than FC18 that I might springboard from?

 We used the olpc-3.10 branch to generate kernels for systems later
 than Fedora 18.

 --
 James Cameron
 http://quozl.linux.org.au/
 ___
 Devel mailing list
 de...@lists.laptop.org
 http://lists.laptop.org/listinfo/devel



 ___
 Server-devel mailing list
 Server-devel@lists.laptop.org
 http://lists.laptop.org/listinfo/server-devel

___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Two (minimal) goals for a kernel for XO1.5 based upon fc22?

2015-05-28 Thread Peter Robinson
On Thu, May 28, 2015 at 9:05 AM, James Cameron qu...@laptop.org wrote:
 The v4.1-rc5 kernel boots fine on XO-1.5, though there are a few
 things to be fixed; screen blanks on boot, camera LED stays on and
 camera doesn't work, temperature of CPU is not accessible, and suspend
 fails to complete.  Some of these probably my haste.

Even more awesome, some of these are probably missing or changed
kernel options. Some of those kernel bits have changed a lot upstream.

 This meets George's requirements.

 The patch relative to v4.1-rc5:
 http://dev.laptop.org/~quozl/y/1YxsfX.txt

Do we have a 4.1 branch in the olpc-kernel git repo yet? Is it worth
it or would it be worthwhile me setting up one against the Fedora
upstream to allow us to more easily work there, rebase etc?

 The boot dmesg:
 http://dev.laptop.org/~quozl/z/1Yxsg9.txt

 You might take your favourite Fedora 22 kernel, apply that patch,
 fixing any conflicts, reconfigure the kernel, and build.

 Or you might compare the xo_1.5_defconfig against the Fedora 22 kernel
 .config.

I'm happy to help there.

 If you guys can help test and fix these apparently minor problems,
 then I can pull upstream into olpc-kernel repository and make a later
 branch.

Can we setup a git branch sooner rather than later as a central place
to pull the latest and greatest from.

 Chances are it would work fine with XO-1 too, with a bit of
 configuration merging.

Happy to help there. We likely need to rebase a few patches for camera
modules etc from the older branches, happy to help there where
possible.

 However, this doesn't help with XO-1.75 and XO-4, which are still just
 as large a job as they ever were.  ;-{  x86 is so easy.

I've got a possible route for the 1.75 as mentioned above, once we get
a 4.1 branch I can add that and see how much further on we can get
with that. XO-4 on the other hand, as you mention, is a massive job!
:-(

Peter

 On Wed, May 27, 2015 at 11:17:26PM -0400, Samuel Greenfeld wrote:
 Given I looked into this before with F22 Beta(*), I ran OLPC OS
 Builder tonight, excluding nothing from all repositories, and let it
 loose.

 There were a few more minor issues found with the build process.
 But I managed to build an image with the stock Fedora 4.x kernel,
 and primarily F22 parts.

 However the stock Fedora kernel does not have a hardcoded kernel
 command line like OLPC kernels do, and trying to manually provide
 the command line to boot a Fedora kernel still hung for me after OFW
 stated the ramdisk was being loaded.

 I may look into this a bit more before resorting back to OLPC's 3.10
 kernel and seeing what works with that.

 (*) [1]http://lists.sugarlabs.org/archive/sugar-devel/2015-April/050035.html

 On Wed, May 27, 2015 at 6:28 PM, James Cameron [2]qu...@laptop.org wrote:

 On Wed, May 27, 2015 at 06:31:28AM -0700, George Hunt wrote:
  If I set just the goals of getting keyboard input, and display output,
 what
  problems will I face trying to use defconfig_xo1.5 from [1][3]
 dev.laptop.org
  (x86-3.3 branch)?
 
  Does anyone have a config file that works on the XO1.5 on a Fedora
 release
  later than FC18 that I might springboard from?

 We used the olpc-3.10 branch to generate kernels for systems later
 than Fedora 18.

 --
 James Cameron
 [4]http://quozl.linux.org.au/
 ___
 Devel mailing list
 [5]de...@lists.laptop.org
 [6]http://lists.laptop.org/listinfo/devel

 References:

 [1] http://lists.sugarlabs.org/archive/sugar-devel/2015-April/050035.html
 [2] mailto:qu...@laptop.org
 [3] http://dev.laptop.org/
 [4] http://quozl.linux.org.au/
 [5] mailto:de...@lists.laptop.org
 [6] http://lists.laptop.org/listinfo/devel

 ___
 Server-devel mailing list
 Server-devel@lists.laptop.org
 http://lists.laptop.org/listinfo/server-devel


 --
 James Cameron
 http://quozl.linux.org.au/
 ___
 Devel mailing list
 de...@lists.laptop.org
 http://lists.laptop.org/listinfo/devel
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Two (minimal) goals for a kernel for XO1.5 based upon fc22?

2015-05-28 Thread Peter Robinson
  Given I looked into this before with F22 Beta(*), I ran OLPC OS Builder
  tonight, excluding nothing from all repositories, and let it loose.

 Funny! I was doing exactly the same yesterday for the 1.5!

  There were a few more minor issues found with the build process.  But I
  managed to build an image with the stock Fedora 4.x kernel, and primarily
  F22 parts.

 Can we create a f22 branch on the o-o-b git repo so we can all
 collaborate on this from one central place?

 Use the master branch, please.  I've moved the Fedora 20 development
 to a v8.0 branch.  So we stick to our existing branch habits.

 http://dev.laptop.org/git/projects/olpc-os-builder

Works for me, just didn't want to step on existing work
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] UEFI workaround?

2014-10-31 Thread Peter Robinson
On Fri, Oct 31, 2014 at 4:50 PM, Adam Holt h...@laptop.org wrote:
 Tony Anderson’s is leaving to Africa in 6 days, assisting many different
 school server deployments, and needs help getting around secure
 BIOS/firmware UEFI on the following platforms:


 - NUC 34010 - BIOS upgrade solves this, by moving jumper and inserting USB
 memory stick containing newer Intel BIOS

 - NUC 2820 - installs but doesn't boot

 - Gigabyte Brix GB-BXBT-2807

 - Zotac nano CI320

 - Zotac ID81 Atom system but UEFI not Bios


 Just in case you have any similar experience, please drop some hints, and
 he'll happily reply here with Tons of more details he's tried :)

Fedora/CentOS should work fine with Secure Boot enabled uEFI devices

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Fwd: [XSCE] UEFI workaround?

2014-10-31 Thread Peter Robinson
So to confirm there is no support for Secure Boot in RHEL-6 and hence
CentOS-6. For this to work you'll definitely need CentOS7.

If the NUC devices are baytrail the issue might not actually be
SecureBoot at all but rather that they have a 32 bit uEFI
implementation and that's not currently supported. Either in 32 bit
CentOS-6 (32 bit doesn't support uEFI at all) and you currently can't
boot a 64 bit OS from a 32 bit uEFI implementation even if the chip is
64 bit.

Peter

 Hi,

 My head is still bloody banging against the brick wall.

 I shipped the NUC 3401 to Nick, he should receive it today. (4gb memory, 1TB
 drive)

 I can reliably create a usb stick from BERNIE to install XS (NEXS v31) on
 systems which have a traditional bios. This scheme does not work on systems
 with UEFI firmware. Microsoft requires
 UEFI firmware for systems certified for Windows 8, so solving this problem
 is unavoidable.

 The Zotac uses AMI firmware while the NUC 2820 and Gigabyte apparently use
 the Intel firmware.
 With the Intel firmware, there is a legacy mode. It also allows the hard
 drive to be installed from the
 usb stick. However, the system will not boot to the hard drive. So the next
 step is to try the boot
 in legacy mode. Another option is to upgrade the bootloader to Grub2 which
 supports uefi. The
 third step is to build an iso which installs a Grub2 bootloader. In the
 longer run, NEXS needs to be
 upgraded to CentOS 6.6 (32-bit) or, more ambitiously, to CentOS 7 (64-bit).

 The Zotac symptom I have also seen with bios systems and I think is really a
 kickstart issue. If
 the install is to a new drive, the usb stick is /dev/sda and the hard drive
 is /dev/sdb. However, if the
 drive has been partitioned (e.g. when I am installing a new version over an
 old one), the hard drive
 is /dev/sda and the usb stick is /dev/sdb. I have created two versions of
 ks.cfg, one for each case.
 However, this is not working for the Zotac which reports that there is not
 enough space on the hard drive instead of re-partitioning.

 Any help will be greatly appreciated.

 Tony



 On 10/31/2014 12:50 PM, Adam Holt wrote:

 Tony Anderson’s is leaving to Africa in 6 days, assisting many different
 school server deployments, and needs help getting around secure
 BIOS/firmware UEFI on the following platforms:


 - NUC 34010 - BIOS upgrade solves this, by moving jumper and inserting USB
 memory stick containing newer Intel BIOS

 - NUC 2820 - installs but doesn't boot

 - Gigabyte Brix GB-BXBT-2807

 - Zotac nano CI320

 - Zotac ID81 Atom system but UEFI not Bios


 Just in case you have any similar experience, please drop some hints, and
 he'll happily reply here with Tons of more details he's tried :)

 --
 Unsung Heroes of OLPC, interviewed live @ http://unleashkids.org !

 --
 Unsung Heroes of OLPC, interviewed live @ http://unleashkids.org !


 ___
 Server-devel mailing list
 Server-devel@lists.laptop.org
 http://lists.laptop.org/listinfo/server-devel

___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] raspi project

2014-10-16 Thread Peter Robinson
On Thu, Oct 16, 2014 at 2:11 PM, Tim Moody t...@timmoody.com wrote:
 I personally have not had much luck with Indiegogo as far as actually
 getting stuff is concerned, but an interesting project if rather expensive.

 https://www.indiegogo.com/projects/pi-top-a-raspberry-pi-laptop-you-build-yourself

It would be even more interesting if it could be coupled with a more
capable Pi compatible device such as the BananaPi or Hummingboard.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] raspi project

2014-10-16 Thread Peter Robinson
 Or something even more powerful, like the iMX6 dual/quad-core EDM module!
 They could at least choose an SoC with published data sheets :-)

That's what the hummingboard is based on I mentioned :
http://www.solid-run.com/products/hummingboard/

 I also love the complete lack of any mechanical testing/specs — I guess nobody
 will ever drop it…

And kids aren't rough wit things at all... ever ;-)

 On Oct 16, 2014, at 9:27 AM, Peter Robinson pbrobin...@gmail.com wrote:

 On Thu, Oct 16, 2014 at 2:11 PM, Tim Moody t...@timmoody.com wrote:
 I personally have not had much luck with Indiegogo as far as actually
 getting stuff is concerned, but an interesting project if rather expensive.

 https://www.indiegogo.com/projects/pi-top-a-raspberry-pi-laptop-you-build-yourself

 It would be even more interesting if it could be coupled with a more
 capable Pi compatible device such as the BananaPi or Hummingboard.

 Peter
 ___
 Server-devel mailing list
 Server-devel@lists.laptop.org
 http://lists.laptop.org/listinfo/server-devel

___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Cubieboard - linux-sunxi

2013-04-21 Thread Peter Robinson
On Sat, Apr 20, 2013 at 11:12 PM, Sameer Verma sve...@sfsu.edu wrote:
 Another ARM board, courtesy of Robert Howard.

 http://linux-sunxi.org/Cubieboard

There's a Fedora 18 remix image that will work with this device available.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Hardware for Schoolservers (EDIT)

2013-01-15 Thread Peter Robinson
On Mon, Jan 14, 2013 at 8:46 PM, German Ruiz germa...@opensuse.org.ni wrote:
 Hello everyone

 Sorry, last message was sent without finish.

 Just asking if anyone here knows about fanless server in the field, in any
 country, i was looking at the wiki[1], and the only information about
 hardware from this vendors [2][3], is in Afghanistan.

 We want to buy a few servers from this companies, to use in Nicaragua, but
 first i want to make sure that this ones works very good.

 Any suggestions?

I've always found the Fit-PC computers quite good as fanless lowpower
computers. The fit-pc 1 actually has the same processor as the
original XO-1

The Fit-PC 2i has 2x Gb ethernet and is low power. They even have AMD
and and Intel core i7 devices with usb3 and dual ethernet as well.

http://www.fit-pc.com/
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Networking issue with XS 0.7 on EPC-AT270

2012-08-01 Thread Peter Robinson
On Wed, Aug 1, 2012 at 2:16 AM, Jerry Vonau jvo...@shaw.ca wrote:
 On Wed, 2012-08-01 at 09:57 +1000, David Leeming wrote:
 Hello,



 I am training some teachers in PNG to set up school servers. We are
 using the EPC-AT270 (brochure attached, specs on page 7) and
 previously have installed X_-v0.6 with no problems at all.



 This time I am trying XS-0.7 but we have a networking issue. It does
 not configure either of the two Ethernet ports.



 The EPC-AT270 has two Ethernet ports.



 The installation is completely default as per the wiki 0.7
 installation guide, from CD, starting with an unformatted HDD. No
 errors are displayed on the screen during installation (but there is
 some scrolling of data).



 After installing, the first symptom is that when plugging in a
 computer or AP to either eth0 or eth1, the port LEDs do not light up
 and no network is seen by the connecting device.



 When querying ifconfig –a I can see eth0 and eth1 but they have no IP
 configuration.



 Any next steps? Meanwhile we are reverting to 0.6 which is OK on the
 EPC

 To late to run mii-tool now, or check which kernel modules were loaded.
 Or check if HWADDR was set in ifcfg-eth0|1 while having 2 nics with the
 same kernel module: http://dev.laptop.org/ticket/11678

mii-tool is mostly (should be completely) obsolete. You should use
ethtool for this functionality as it will work with all detected NICs.
dmesg | grep eth will let you know what the kernel sees.

The Marvell ethernet ports might not be well supported on CentOS 6.
You might want to check if a newer CentOS kernel adds support.
Alternatively you can likely use a Fedora kernel. dsd did that when he
had issues with NIC support on one of the deployments he was involved
with. I suggest checking the mailing list archives for the details.

Regards,
Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Networking issue with XS 0.7 on EPC-AT270

2012-08-01 Thread Peter Robinson
On Wed, Aug 1, 2012 at 8:52 PM, David Leeming
da...@leeming-consulting.com wrote:
 Thanks Peter,

 dmesg | grep eth

 shows that the kernel sees the hardware, at least eth0 and eth1 hardware/MAC
 addresses are shown, and those agree with that shown from ifconfig -a which
 shows eth0, eth1 and lo but only lo with inet address (loopbackl). The
 physical ports appear to be inactive, the LEDs don't light up and when
 trying with a cable and computer, no network detected.

Does ethtool eth0 report anything? And lspci|grep Ethernet What
kernel are you using? Can you provide more details?

Peter

 Reinstalling with 0.6, the networking works fine

 David

 -Original Message-
 From: Peter Robinson [mailto:pbrobin...@gmail.com]
 Sent: Wednesday, 1 August 2012 7:28 p.m.
 To: Jerry Vonau
 Cc: David Leeming; XS Devel
 Subject: Re: [Server-devel] Networking issue with XS 0.7 on EPC-AT270

 On Wed, Aug 1, 2012 at 2:16 AM, Jerry Vonau jvo...@shaw.ca wrote:
 On Wed, 2012-08-01 at 09:57 +1000, David Leeming wrote:
 Hello,



 I am training some teachers in PNG to set up school servers. We are
 using the EPC-AT270 (brochure attached, specs on page 7) and
 previously have installed X_-v0.6 with no problems at all.



 This time I am trying XS-0.7 but we have a networking issue. It does
 not configure either of the two Ethernet ports.



 The EPC-AT270 has two Ethernet ports.



 The installation is completely default as per the wiki 0.7
 installation guide, from CD, starting with an unformatted HDD. No
 errors are displayed on the screen during installation (but there is
 some scrolling of data).



 After installing, the first symptom is that when plugging in a
 computer or AP to either eth0 or eth1, the port LEDs do not light up
 and no network is seen by the connecting device.



 When querying ifconfig -a I can see eth0 and eth1 but they have no IP
 configuration.



 Any next steps? Meanwhile we are reverting to 0.6 which is OK on the
 EPC

 To late to run mii-tool now, or check which kernel modules were loaded.
 Or check if HWADDR was set in ifcfg-eth0|1 while having 2 nics with the
 same kernel module: http://dev.laptop.org/ticket/11678

 mii-tool is mostly (should be completely) obsolete. You should use
 ethtool for this functionality as it will work with all detected NICs.
 dmesg | grep eth will let you know what the kernel sees.

 The Marvell ethernet ports might not be well supported on CentOS 6.
 You might want to check if a newer CentOS kernel adds support.
 Alternatively you can likely use a Fedora kernel. dsd did that when he
 had issues with NIC support on one of the deployments he was involved
 with. I suggest checking the mailing list archives for the details.

 Regards,
 Peter



___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] armv7hl vs armv7l

2012-07-06 Thread Peter Robinson
On Fri, Jul 6, 2012 at 10:42 AM, George Hunt georgejh...@gmail.com wrote:
 Hi Peter,

 You probably know the answer to this question off the top of your head.

 I've played with fedora's Trimslice armv7hl, using it to recompile XS rpms.
 Now in conversation with OLPC-Australia, I've agreed to try to apply my
 stuff to the XO-1.75 pre-release 12.1.0, which I believe is based upon FC17.

It is indeed, it's using the F-17 arm hardfp release.

 Question: Is my easiest path to basically start over, either building up a
 cross compiling tool chain, or maybe try to compile the XS rpms on an armv7l
 machine natively, as I did with the TS, (the XO itself seems the obvious
 choice).

If you have a trimslice why don't you use that and compile natively?
In Fedora everything is compiled natively with no cross-compilation.

 I had trouble earlier getting a tool chain together to run on FC17, on top
 if parallels, on my MAC.

To be honest I've never cross compiled any ARM packages.

 Do you have any advice?

Compile natively :-)

If you have a Trimslice, Pandaboard or even an XO 1.75 you can compile
on all of those using the standard distros. On any of the platforms
you can yum install or yum groupinstall  anything you may need and
build directly. You might want to add an ext4 formatted usb HDD to use
as the storage for building on those platforms as they tend to be a
bit quicker than SD card storage.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] armv7hl vs armv7l

2012-07-06 Thread Peter Robinson
On Fri, Jul 6, 2012 at 11:52 AM, George Hunt georgejh...@gmail.com wrote:
 Thanks Peter,

 I was confused when I installed latest 12.1.0 on an XO and issued uname
 -a, to see the response come back armv7l, rather than armv7hl.  I was
 thinking that yum would be confused by the difference.

The kernel isn't hardfp/softfp because the kernel doesn't use maths
co-processors. rpm/yum has been hacked to deal with it. It's all a
little ugly but it works.

 I'm glad that the trimslice generated rpms I have will be usable.  I'll need
 to learn how to override the default arch, so that yum will do what I want
 it to do. But I have google for that!

Why do you need to override the arch? What exactly are you trying to do?

 Thanks for your help,

 George


 On Fri, Jul 6, 2012 at 6:39 AM, Peter Robinson pbrobin...@gmail.com wrote:

 On Fri, Jul 6, 2012 at 10:42 AM, George Hunt georgejh...@gmail.com
 wrote:
  Hi Peter,
 
  You probably know the answer to this question off the top of your head.
 
  I've played with fedora's Trimslice armv7hl, using it to recompile XS
  rpms.
  Now in conversation with OLPC-Australia, I've agreed to try to apply my
  stuff to the XO-1.75 pre-release 12.1.0, which I believe is based upon
  FC17.

 It is indeed, it's using the F-17 arm hardfp release.

  Question: Is my easiest path to basically start over, either building up
  a
  cross compiling tool chain, or maybe try to compile the XS rpms on an
  armv7l
  machine natively, as I did with the TS, (the XO itself seems the obvious
  choice).

 If you have a trimslice why don't you use that and compile natively?
 In Fedora everything is compiled natively with no cross-compilation.

  I had trouble earlier getting a tool chain together to run on FC17, on
  top
  if parallels, on my MAC.

 To be honest I've never cross compiled any ARM packages.

  Do you have any advice?

 Compile natively :-)

 If you have a Trimslice, Pandaboard or even an XO 1.75 you can compile
 on all of those using the standard distros. On any of the platforms
 you can yum install or yum groupinstall  anything you may need and
 build directly. You might want to add an ext4 formatted usb HDD to use
 as the storage for building on those platforms as they tend to be a
 bit quicker than SD card storage.

 Peter


___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] ARM on XS -- how can I integrate my work?

2012-06-28 Thread Peter Robinson
On Thu, Jun 28, 2012 at 6:11 AM, rihowa...@gmail.com
rihowa...@gmail.com wrote:


 On Jun 25, 2012, at 8:27 AM, Daniel Drake wrote:

 On Sun, Jun 24, 2012 at 8:14 AM, George Hunt georgejh...@gmail.com wrote:
 I'm not done yet, but I've been making progress on porting XS code to ARM by
 making modifications to DSD's  XS-0.7.  Upon his suggestion, I have been
 basing my work on the srpms posted at
 http://xs-dev.laptop.org/xsrepos/stable/olpc/xs-0.7/source/.

 Now that I've got some of the services running, I'm wondering how to
 contribute to the XS codebase.  What I'd prefer is to contribute deltas from
 XS-0.7 that use `uname -p` to enable the appropriate path through the
 startup scripts.

 That kind of approach would suggest supporting both CentOS and F17+.
 I'm not sure if thats the direction we'd want to go - supporting 2
 platforms has its costs. It might be preferred to do a full migration
 to F17. You'll need clarification from this from Martin, who's away
 until next month.

 There is no CentOS for ARM.  It is not even on the horizon yet. On the other
 hand rpmfusion has recently expressed interest in building for the 2 ARM
 archs supported by Fedora.

There's no CentOS for ARM because there is no EL for ARM, there's is a
project attempting to achieve a EL6 clone but IMO it's a waste of time
as it can only support armv5tel because the compiler and tool chain in
EL6 doesn't even know hardfp exists.

rpmfusion was planning on adding ARM support when they move to koji
from their previous (current?) build platform. Not sure of the current
status of that.

 I have been targeting F17 for ARM with an eye to F18 for ARM which may
 make the cut to become a primary architecture.

It's unlikely at this point in time that ARM will become primary arch
in the F-18 timeframe simply due to one of the big blockers is decent
build hardware and while it should be available soon it's not going
to be enough time to deal with that. At this point in time F-19 is
much more likely.

Ultimately any packages or changes needed for XS whether it's aimed at
CentOS/EPEL or Fedora whether it be on x86 or ARM should aim to
getting the changes into mainline Fedora which means if it's in
mainline it will be build on ARM whether ARM is primary or secondary
arch. It generally makes the maintenance of the packages easier and
more straight forward especially if it's a change/enhancement to
existing packages as the changes evolve when the main package evolves.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] School server on ARM related question: Systemd on fc17 doesn't see ethernet dongles or USB sticks

2012-05-20 Thread Peter Robinson
On May 19, 2012 8:22 PM, George Hunt georgejh...@gmail.com wrote:

 Hi all,

 There are no entries in /var/log/messages or dmesg relating to my test
USB stick at /dev/sdb1, even though during the boot process, dracut sees an
8 GB sdb as well as the 132GB hard disk on sda that is the rootfs (so I
think the proper drivers are in the kernel). I'm playing with the trimslice
root fs and kernel snapshots available at
http://fedoraproject.org/wiki/Architectures/ARM.

The USB issue was a timeout issue and was fixed as of kernel 3.3.4-5 so
make sure you're running that, else its likely something else.

 I also cannot configure an ethernet dongle, which the kernel doesn't
appear to respond to, when inserted.

Does it work on a f17 x86 device?

 Any pointers to documentation, or what modules's source I might study? I
don't really understand how HAL was replaced, or why there are no entries
in the kernel log.

Firstly you're better asking this on the fedora arm list as it's better
suited to generic arm hardware problems.

For the messages do they appear when you run dmesg? If so you need to
enable syslog. systemctl enable syslogd.service from memory and reboot.

As for the dongle it could be many and varied. If it's supported and all
the modules are there it should just work. Test it on am x86 box .

Peter

 Thanks,

 George

 ___
 Server-devel mailing list
 Server-devel@lists.laptop.org
 http://lists.laptop.org/listinfo/server-devel

___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS on ARM update.

2012-04-12 Thread Peter Robinson
On Thu, Apr 12, 2012 at 7:12 AM, rihowa...@gmail.com
rihowa...@gmail.com wrote:



 Note that currently Puppet relies on some hard coded intel specific code.

 Oh really? I'm assuming it'd be in the facter code, using lspci and
 dmidecode to get the facts about hardware

 I just checked ARM Koji and it looks like puppet is building okay on ARM and 
 passing the tests on ARM.
 I must be remembering some problem it had during the fedora 15 moji phase, 
 but then
 a number of things had problems then.

F15 was a mess, it was used solely as a bringup mechanism for hard
float. F17 is the only way to go and if you've got any issues let me
know as I've pretty much built it all.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Looking for new low power server hardware candidate

2012-04-12 Thread Peter Robinson
On Wed, Apr 11, 2012 at 4:58 PM, rihowa...@gmail.com
rihowa...@gmail.com wrote:


 On Apr 10, 2012, at 5:04 PM, Martin Langhoff wrote:

 On Mon, Apr 2, 2012 at 2:03 AM, George Hunt georgejh...@gmail.com wrote:
 I left my fitpc2 and msi servers in the Philippines, hoping they would be
 pressed into service in a classroom situation.  So now I'm in the market for
 another toy.

 If you have time towork with us through some hitches, I'd recommend an
 ARM server. At this stage I'd say one of the Marvell/Globalscale
 Plug servers (dreamplug for example), or a trimslice.

 Either option will need a combination of the OS on internal SD/eMMC
 and the storage on an ext HDD (via USB probably).

 The Kirkwood based systems such plug computers can boot both the kernel and 
 OS from a hard
 drive. I have been talking with one of he Fedora ARM team about this and am 
 going
 to send them an email about what is involved and some other things about the 
 kirkwood based devices.

I'm one of the Fedora ARM people so feel free to ask here, there's a
lot of work going on to simplify the process of creating images, there
should be some more stuff coming soon for creation of images. I would
suggest looking at ARMv7 devices as they tend to have more CPU power
and memory which would be better for the server stuff. Personally I
think the Trimslice H is going to one of the best models as you can
put a decent HDD in there and have a self contained unit, although I
would love a dual eth option.

Peter

[1] http://trimslice.com/web/models
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Looking for new low power server hardware candidate

2012-04-11 Thread Peter Robinson
On Wed, Apr 11, 2012 at 1:53 PM, Martin Langhoff
martin.langh...@gmail.com wrote:
 On Tue, Apr 10, 2012 at 10:55 PM, John Watlington w...@laptop.org wrote:
 Why not an XO-1.75 ?

 Good point. And XO + AP + HDD would work fantastic.

 George, how many users per server? If 100, an XO-1.75 will do ok.
 Want to sign up for the Contributors Programme (search in the wiki for
 the URL).

 XO-1.75, Plug or Trimslice will do fine with a recent Fedora for ARM
 (from the upcoming F17 series) -- we just need to recompile the XS
 specific packages. Most of them will just work. AFAIK, ejabberd and
 xs-config will need some work, and I can help you with those.

It might be worthwhile seeing what we can get into mainline Fedora for
the XS releases so that as the new RHEL releases come along it will
just work as well as be usable on ARM platforms.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] initial notes on 0.7

2012-02-15 Thread Peter Robinson
On Thu, Feb 16, 2012 at 12:43 AM, Sameer Verma sve...@sfsu.edu wrote:
 On Sun, Feb 12, 2012 at 12:18 AM, Sameer Verma sve...@sfsu.edu wrote:
 On Sat, Feb 11, 2012 at 2:17 PM, Sameer Verma sve...@sfsu.edu wrote:
 We managed to install XS 0.7 using the CentOS 6.2(minimal) + EPEL +
 OLPC XS approach. The machine is SolidLogic box (same as the one used
 in OLPCorps). The installation went well, with minor glitches (notes
 will be on the wiki).

 One pattern we've seen is that on XO 1.5 build 883, it keeps
 restarting X (gives some dcon freeze message) periodically. On a XO
 1.5 build 852, it runs ok. Restarting problem was on three different
 XO 1.5s.

 cheers,
 Sameer
 --
 Sameer Verma, Ph.D.
 Professor, Information Systems
 San Francisco State University
 http://verma.sfsu.edu/
 http://commons.sfsu.edu/
 http://olpcsf.org/


 Some more details based on notes (thx to Alex Kleider).

 Machine used: SolidLogic GS-L02 Fanless Mini-ITX System
 http://www.logicsupply.com/products/systemgs_l02

 Note: We tried to install on a Fujitsu P2120
 (http://www.linuxsoftware.co.nz/wiki/P2120Linuxr) but it failed
 because CentOS 6.2 won't install on non-PAE systems. This machine was
 able to run XS 0.6


 What distro would be appropriate to start for a non-PAE system?

You should be able to use the above with a non Fedora 15 non PAE kernel.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS-0.7 CentOS6.2 rebase - other pending items

2012-02-10 Thread Peter Robinson
On Thu, Feb 9, 2012 at 10:48 PM, Daniel Drake d...@laptop.org wrote:
 Hi,

 I now have an XS fully up and running and passing all my basic tests.
 Here are the remaining items that need addressing before we have a
 test release:

 ejabberd - see the other thread. Need to decide on forking the package
 as 'ejabberd' or 'ejabberd-xs' to move forward. Once that is done, I
 will update xs-config (if needed) and then push and release all the
 other components you have already reviewed.

 moodle - pu branch ready for review. If you're going to pull in moodle
 updates as well, now is the time :)
 I have tested this quite well, including the interaction with mod_admin_extra.

 xs-release - how do we go forward with this? I think we should drop
 the old approach (of *replacing* the system release package) and take
 the epel-release approach of just (additionally) installing our repo
 files. But I'm not sure how you want this in git - existing branch of
 existing repo, new repo? Or maybe I could create a new
 packages/xs-release repo, with all the files contained in the spec
 file repo (i.e. doesn't pull in a tarball, just ships the trivial repo
 files directly).

 xs-logos - Haven't really looked what this has. Given that we don't
 face copyright/trademark restrictions of the logo package in CentOS,
 can we just drop this?

 usbmount - I had to update to the latest version. It no longer uses
 any patches (they are all obsolete/upstream). How do I take care of
 this w.r.t. your existing usbmount git repository, where you actually
 forked the source? Perhaps we could just drop/obsolete that git repo,
 and create a new packages/usbmount repo with the simple .spec file?

 olpc-xs-builder - pu branch ready for review. Might need tweaking
 based on the outcome of the above. I dropped the idea of running
 xs-setup during the install, since the user might choose a hostname
 that doesn't start with schoolserver.. The installation instructions
 will require the user to run xs-setup after the install completes.

 repos - I have reorganised slightly http://dev.laptop.org/xs/
 repos is now a subdirectory there, which will be our main URL from now on.
 But the other URLs still work: http://dev.laptop.org/xsrepos/
 http://dev.laptop.org/~martin/xsrepos
 Also, I have created aliases at http://dev.laptop.org/xs/stable and
 http://dev.laptop.org/xs/testing for the repos. This means that if we
 update the DNS of fedora.laptop.org, we will fix yum update / yum
 install for the existing XS's in the field, which use such addresses.
 What do you think?


 I had to bring some packages in from Fedora, these are:

 bitfrost-1.0.15-3.el6.i686.rpm - not in RHEL/EPEL. Recompiled for EL6
 from rawhide.
 mtd-utils-1.3.1-3.fc14.i686.rpm - dep of bios-crypto, imported from F14

We can get branches created for EPEL for anything missing if that
would help, let me know if you want help.

 kernel-2.6.42.2-1.fc15.i686.rpm - as previously agreed, imported from F15
 (kernel-* subpackages too)
 grubby-7.0.16-5.fc15.i686.rpm - dep of kernel, imported from F15
 linux-firmware-20110601-1.fc15.noarch.rpm - dep of kernel, imported from F15
 module-init-tools-3.16-2.fc15.i686.rpm - dep of kernel, imported from F15
 acpid-2.0.9-1.fc14.i686.rpm - imported from F14. Needed for compat
 with new kernel.

 rssh-2.3.3-2.el6.i686.rpm - imported from EPEL-6 updates
 syck-python-0.61-12.el6.i686.rpm - dep of ds-backup, not in RHEL/EPEL.
 F14 version recompiled for EL6.
 syck-0.61-12.el6.i686.rpm - dep of syck-python

 Is it OK to stick these in the core xs-0.7 RPM repo, or would you
 prefer a separate fedora-ports repo to be created? (I vote just the
 one :))


 Thanks,
 Daniel
 ___
 Server-devel mailing list
 Server-devel@lists.laptop.org
 http://lists.laptop.org/listinfo/server-devel
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS rebase review

2012-02-08 Thread Peter Robinson
On Wed, Feb 8, 2012 at 4:27 PM, rihowa...@gmail.com rihowa...@gmail.com wrote:


 On Feb 8, 2012, at 3:35 AM, Martin Langhoff wrote:

 On Wed, Feb 8, 2012 at 2:23 AM, rihowa...@gmail.com rihowa...@gmail.com 
 wrote:
 Which releases of erlang and ejabbered are you using?

 I unfortunately don't have the version number handy, but this should
 help: this coming XS release only ships a configuration and init
 scripts for ejabberd -- it's based on RHEL6.2/CentOS6.2, using the
 ejabberd versions included in the OS.

 I took a look at the CentOS6.2 source repository at 
 http://vault.centos.org/6.2/ and could not find Erlang or ejabberd there.

It's in EPEL

https://fedoraproject.org/wiki/EPEL

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] CentOS hardware support doubts

2012-02-03 Thread Peter Robinson
On Fri, Feb 3, 2012 at 2:07 PM, Daniel Drake d...@laptop.org wrote:
 On Thu, Feb 2, 2012 at 9:09 PM, Martin Langhoff
 martin.langh...@gmail.com wrote:
 I assume here that CentOS is reasonably in sync with RHEL. Does
 http://elrepo.org/bugs/print_bug_page.php?bug_id=126 help? More
 generally, does any of the external repos have a kmod-staging or
 kmod-atl1e that works for you?

 I've returned that system now; if I get the time and opportunity to
 test again, I will do so. Where is the list of external repos?

 I'm worried about the expertise required in order to identify such
 repos and packages. We need this process to be doable without me in
 the room.

 My assumption is that RHEL/CentOS have fairly decent hardware support
 from backported drivers, some in the RH kernels, EPEL or external
 repos (in order of decreasing quality expectations...).

 I assume that RHEL is pretty good for server-class hardware found in
 US/EU; I can imagine why the support of desktop-class hardware found
 in the poorer parts of latin america may be lesser so.

 I wonder if you've been unlucky in the mix of hw you got there; or
 whether the driver support situation for essential things like NICs
 and disk controllers is weaker than I had expected. Maybe others with
 more practical experience with current RHEL/CentOS can comment...?

 I've now seen 3 failure cases - the AR8152 mentioned above, and
 another case which I only had time to do a quick boot check of
 F9/C6/F16 (F16 was the only one that recognised the onboard NIC of the
 asrock motherboard).

 Yesterday we received 10 servers based on an Intel motherboard (and 12
 more will be coming next week). F9 doesn't recognise the onboard NIC.
 C6 recognises the onboard NIC but isn't able to send/receive packets.
 F16 works fine (using e1000e driver). As these boards only have 1 PCI
 socket it is not possible to have 2 NICs (unless we resort to USB...)
 unless we move beyond C6.
 Also, F9 and C6 do not recognise the SATA DVD drive in these systems -
 no /dev/sr0 created, error in dmesg during boot. This will be a pain
 for field work. With F16 this works fine.

 I haven't yet found a case where the F9--C6 upgrade adds hardware
 support for any hardware that we have here.


 I like your idea of using a F16 kernel on top of CentOS 6.2. So far,
 his seems to be working fine (and solves all of the compatibility
 problems mentioned above). If this continues to work I would like to
 push it as the default for XS install media.

I suggest using the F-15 kernel. The 2.6.42.x kernel in F-15 is the
3.2.x kernel but there's issues with a number of utilises plain not
working because they can't work out the kernel version because they
don't expect a major version of 3.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] CentOS hardware support doubts

2012-02-03 Thread Peter Robinson
On Fri, Feb 3, 2012 at 2:38 PM, Samuel Greenfeld greenf...@laptop.org wrote:

 On Fri, Feb 3, 2012 at 9:07 AM, Daniel Drake d...@laptop.org wrote:

 I've now seen 3 failure cases - the AR8152 mentioned above, and
 another case which I only had time to do a quick boot check of
 F9/C6/F16 (F16 was the only one that recognised the onboard NIC of the
 asrock motherboard).

 Yesterday we received 10 servers based on an Intel motherboard (and 12
 more will be coming next week). F9 doesn't recognise the onboard NIC.
 C6 recognises the onboard NIC but isn't able to send/receive packets.
 F16 works fine (using e1000e driver). As these boards only have 1 PCI
 socket it is not possible to have 2 NICs (unless we resort to USB...)
 unless we move beyond C6.

What's the exact version of the centos kernel? There always seems to
be lots of new revisions of the e1000 cards that need slightly newer
drivers. Have you checked out the Centos Continuous release repo to
see if a newer kernel is available?

http://lists.centos.org/pipermail/centos-announce/2011-September/018078.html

 It's worth noting that if you have to, there are NIC cards available with
 more than one port per PCI slot.  They just tend to be rarer and as
 server-grade hardware, more expensive.

No so much now. I think our cost price for 2 port cards is around
£30-40 for standard cards. Depending on the switch it's being attached
to you could also use vlan trunking.

 Coming from a networking/ODM background, I have worked with plenty of 2-8
 port e1000 NIC cards, and even 8-port tulip adapters.  Just make sure that
 the PCI-E/X/etc. slot you are using has enough lanes to fit the NIC card in
 the slot, and for the load a schoolserver generates you should be fine.

PCI-e 1.0 is 1Gb per lane. PCI-e 2 is double. On newer boxes it should
be the later so a single lane is generally enough for a 2x 1Gb card.

 Historically I have seen e1000's and Broadcom Gigabit adapters in
 server-grade hardware.  But given I have been out of the industry for a few
 years, I don't know what companies are using nowadays.

Nothing has changed there what so ever at the 1Gb level :-)

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Who wrote http://wiki.laptop.org/go/XS_Install_Server?

2012-01-31 Thread Peter Robinson
On Tue, Jan 31, 2012 at 3:20 PM, Daniel Drake d...@laptop.org wrote:
 On Mon, Jan 30, 2012 at 11:40 AM, George Hunt georgejh...@gmail.com wrote:
 I met Tony Anderson in Haiti, and again at the San Francisco OLPC Summit in
 late 2011. He prevailed upon me to spend some time trying to figure out how
 to rebase XS on a more recent Fedora Core.

 Yesterday I also started looking at that task. Martin explained that
 he'd like to see it based on Centos 6.2 and installable as a group of
 packages on top of a vanilla base install. Hopefully he will send a
 few more details soon (he's travelling).

That would be great. Even better if the packages ended up in EPEL so
that users could install it on which ever of the 4-5 variants of EL-6
they preferred to use. Let me know if you need assistance in this
route.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] using a USB modem for Internet access

2011-10-08 Thread Peter Robinson
On Sat, Oct 8, 2011 at 8:00 AM, Sameer Verma sve...@sfsu.edu wrote:
 I have a XS setup at a remote location in India, where the XS has a
 Ethernet port set up as WAN, and a Mesh antenna (USB) set up as the
 LAN interface. We've used the Ethernet WAN port via a crossover cable
 via Internet-connection-sharing via Windows XP via a USB 3G modem
 (make sense?). I use OpenVPN to establish a tunnel to this machine
 whenever it gets online. This may sound like a complicated setup, but
 it works, except for the Windows XP part, which is a bit of work every
 time and not very reliable.

 Now, we have the possibility of using a compatible USB 3G modem to get
 to the Internet directly. I am looking at
 http://wiki.laptop.org/go/XS_Techniques_and_Configuration#Using_a_different_WAN_connection
 to setup the USB stick as the WAN port. I have one shot at this.
 Essentially, when I get to the XS, it will be through the Ethernet
 port, but when I change things over, its a one shot deal. Anything
 else I should concern myself with? Any gotchas?

The gnome internet sharing should happily share your 3G connection via
wired or wireless quite successfully if you have the gnome UX
installed.

Cheers,
Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS on the XO 1.5

2011-08-16 Thread Peter Robinson
On Tue, Aug 16, 2011 at 3:10 AM, Sameer Verma sve...@sfsu.edu wrote:
 I was thinking (once again) about the possibility of running XS on a
 XO 1.5. On the XO-1 the built-in radio runs in the 802.11s mesh mode
 and serves out IPs via DHCP. Given that the 1.5 does not do 802.11s
 mesh, can't the radio instead simply work in adhoc mode and serve the
 same purpose? I am not sure what the limitations would be of adhoc vs
 mesh mode in terms of number of clients.

I don't believe adhoc uses gateways which would likely cause issues,
but from gnome you should be able to share the wifi and use it as a AP
which should issue IPs, GW etc.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] [Sugar-devel] [Dextrose] Sugar Server project initiation announce

2011-06-11 Thread Peter Robinson
Again, I like where this discussion is going, so it may be worthwhile
 to take some of this back to the drawing board. There is the issue of:

 1) distro independence


If you want distro independence you end up with a document of best practices
on how to install a XS/Sugar server. That's not a server. If you think that
its easy to produce a working product you are wrong. This isn't a single
service such as a DB.This is a collection of services that interact with
each other and have dependencies and other interactions. You can either end
up with a single product running on a single distro, or you can have a
document documenting each of the components and then map them to various
distros. Its a lot of engineering time and resources to do both. One which I
would doubt AC have the interest in doing and a single person working on it
for 6 months certainly won't have the time to do!

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] [OLPC-AU] OLPC Australia XS concerns

2011-06-10 Thread Peter Robinson
On Fri, Jun 10, 2011 at 1:54 AM, Sridhar Dhanapalan
srid...@laptop.org.au wrote:
 I think my biggest technical concerns in XS-land are twofold:

  * we need the XS to behave well on an *existing* network (i.e.
 single interface), without trying to be a gateway or duplicating core
 network services (DNS, DHCP, etc.)
  * while other XS efforts are keen to add features, we want to be as
 trim as possible

 We started the XS-AU when it had become clear that XS development had
 slowed. We could find no alternative that satisfied our needs, and I
 felt it better to go our own way rather than complaining that the XS
 didn't meet our particular use case (which seems to be quite different
 from other deployments). It's been working very well, and it's quite
 low-maintenance for us. We'll need a good reason to jump ship.

 We've been working on a prototype XS Lite, which is essentially an
 XS-AU with everything except ejabberd stripped away. Our deployments
 are done at the classroom-level; a teacher receives XOs for the
 children in their class once they have completed the necessary
 training. We would like to provide a simple server with that
 allocation of XOs. This means that the server needs to be low-cost and
 easy to implement (plug-and-play). We are assuming that there is *no*
 technical expertise available at the school.

 The server doesn't have to be very capable. Anything that requires
 registration won't work for us as the turnover of teachers and
 students is too high. We don't need Moodle or anything similar, since
 such services are already provided on the state education network.
 Since it's based on the XS-AU, it can be 'upgraded' to a full XS with
 some yum commands.

 Given the modest requirements, I think an XO would be suitable
 hardware. They are cheap and reliable, and we already have them in
 stock. As a standalone collaboration server, the XO's WLAN can be the
 AP. If we need to connect to the school network, we can use a
 USB2Ethernet adapter. This will also allow the server to leverage the
 other APs in the school. What's important is that we need to be
 tolerant of multiple schoolservers on the network, potentially one per
 class.

To be honest I think your actually going about it all wrong. XS is
designed to be a do it all server for people that don't want to think.
If you only want a jabber server (or certain components) I think you'd
be much better off just using CentOS and installing ejabberd. Why
start big and remove everything when its easier just to start with a
base server and add one thing? You could even create a virtual
appliance to plug into any virtual infra that the school might have,
boot it off a USB key or what ever.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Observations and question about http://wiki.laptop.org/go/XS_Software_Repositories

2010-09-27 Thread Peter Robinson
On Mon, Sep 27, 2010 at 2:56 PM, Martin Langhoff
martin.langh...@gmail.com wrote:
 On Mon, Sep 27, 2010 at 1:58 AM, rihowa...@gmail.com
 rihowa...@gmail.com wrote:
 I am back to working on getting the XS working on one of my ARM systems with 
 a more recent Fedora version.

 Hi Robert

 - that's excellent news! Ask on de...@lists.laptop.org -- AIUI, there
 is work afoot for F13 or F14 on ARM. I don't really know whether there
 are any usable builds...

There's more details about ARM on Fedora on the project page (which
also links to the arm mailing list etc).

http://fedoraproject.org/wiki/Architectures/ARM

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Observations and question about http://wiki.laptop.org/go/XS_Software_Repositories

2010-09-27 Thread Peter Robinson
On Mon, Sep 27, 2010 at 3:01 PM, Martin Langhoff
martin.langh...@gmail.com wrote:
 On Mon, Sep 27, 2010 at 9:59 AM, Peter Robinson pbrobin...@gmail.com wrote:
 There's more details about ARM on Fedora on the project page (which
 also links to the arm mailing list etc).

 http://fedoraproject.org/wiki/Architectures/ARM

 Yeah, but I looked at it earlier today and it still touts the F-12
 build as the latest.

 The list archive shows someone asking last month about F-13 images,
 and getting no answer.

 So to an outsider it looks like F-12 is the latest. Is that so?

They've had to setup a completely new koji infra as the people that
were running the arm build up to F-12 changed. The F-13 and later
builds are partially complete but there's been issues getting enough
systems to be able to fully build and also some issues with the build
devices they have wrt to power/network issues. OLPC added 3 or 4
builders to the koji instance to help improve the situation. I believe
all the core and lot of the other bits are complete.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] problems installing xs-activation on F12

2010-04-27 Thread Peter Robinson
On Tue, Apr 27, 2010 at 1:44 PM, Daniel Drake d...@laptop.org wrote:
 http://fedora.laptop.org/xs/stable/olpc/xs-0.6/i386/xs-activation-0.2.39.g2277cdf-1.xs9.noarch.rpm

 Straightforward rpm -ivh of the RPM gives dependency errors. It needs:
 olpc-contents
 python = 2.5
 python-json
 usbmount
 xs-tools


 olpc-contents and python-json easily installed by yum.

 xs-tools requires python 2.5.

 usbmount installs with the RPM.

 The packages that need python-2.5 install their files in the wrong
 place for a python-2.6 system.

Does a recompile of this against a F-12 system not fix the python 2.5
- 2.6 problem?

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS on Mac Mini

2010-03-02 Thread Peter Robinson
On Tue, Mar 2, 2010 at 6:19 AM, David Leeming
da...@leeming-consulting.comwrote:

  Sorry if this has been asked before. The Mac Mini was tested by someone
 here with a watt meter and ran between 16 and 30 watts max. Very useful for
 solar powered locations.



 Are there any issues with installing XS on one?

I don't believe there should be any issues as the Mac Minis are well
supported on Fedora which is what the XS is based upon. The Mac Mini servers
with 2 disks would be good as they provide some resilience.

The other device you may want to look at for low powered devices are the
eeeBox devices which are atom based so will use even less power.

Regards,
Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] We need: examples of XS hardware in use --

2010-01-12 Thread Peter Robinson
On Tue, Jan 12, 2010 at 1:00 PM, Hamilton Chua hamilton.c...@gmail.com wrote:
 Hello Martin,

 Happy New Year !

 We run our XS inside virtual machines. I have XS inside virtualbox on my
 Lenovo T61 laptop

 Santa Rosa Core 2 DUO 2.0 Ghz
 4GB DDR2
 120GB HDD

 while Solutiongrove runs a couple of XS installations inside VMWare ESXi
 (http://www.vmware.com/products/esxi/).

 I'll try to get the hardware info where ESXi is installed from our
 service provider.

 ESXi has these fancy charts and graphs on usage and stuff about the
 virtual machines running, I'll see if it's possible to extract that info
 for you. Hopefully that data is still available.

In the case of VMs you need the resources allocated to the VMs rather
than the hardware specs as the hardware specs don't really mean much.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Early F11 XS build

2009-10-24 Thread Peter Robinson
Hi Martin,

On Fri, Oct 23, 2009 at 9:48 PM, Martin Langhoff
martin.langh...@gmail.com wrote:
 In between various complications, I have been working on getting an
 initial rebase of the XS packages and build infra to F11.

 After much wrangling with revisor, comps and image-creator, I have an
 initial installer iso and an image that boots on the XO-1. The XO
 image tests reasonably well, I haven't had chance to play much with
 the installer iso (but the image is built from it).

 All of the above owes a lot to hints from Jerry Vonau - thanks!

 You will find the files on http://xs-dev.laptop.org/xs/other/ in
 about... 6hs -- the name says 0.7 which is misleading, so I will rm
 them once I have better ones. But in case you want something to play
 with :-)

 Notes:

  - The XO image is using the old XO kernel  --
   -- I could not include the new kernel from dillinger's repo due to
 a conflict with kernel-firmware, even if say -kernel-firmware in the
 ks.
   -- the vanilla F11 kernels fail to boot on the XO -- expected?

I think it should boot but I don't think the boot.fth is created but
its been a while since I've got around to trying it on my XO.

  - The build is slimmer (without a need to slim-it-down manually)
 because upstream now avoid including some if the boot images in the
 main iso.

  - However, that means that making a bootable usb is harder for users.

What causes it to be harder to make a bootable usb?

  - Dependency chains are slimmer... but some are still bizarre. Bits
 of kde and gnome are there.

At a guess this is probably due to PolicyKit. I've had issues with it
in the past.

 Digests and sizes

 546M    /srv/revisor/xs-f11-i386/iso/OLPC School Server-0.7-i386.iso
 401M    OLPC School Server-0.7-i386.img.gz

 aa41b6dbca1aac13f842f13f8106019276d065e6  OLPC School Server-0.7-i386.iso
 6bfe1a507704e21e84660a8d091624ffc7dee17a  OLPC School Server-0.7-i386.img.gz

Cheers,
Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS - rebase on F-11, notes and RHF

2009-10-08 Thread Peter Robinson
Hi Martin,

0.6 looks like a great release.

On Thu, Oct 8, 2009 at 10:48 AM, Martin Langhoff
martin.langh...@gmail.com wrote:
 So the F-9-based XS-0.6 is closed (at long last), and the next step is
 to prep 0.7 based on F-11. Here is a rundown of the packaging /
 installer work I think needs to be done:

 =Rebuild packages, spec fixups=

 Jerry Vonau has done a lot of the exploratory work, so mostly done I
 think. Thanks! Still, might hit issues and cry for help...

 =Switch from metapackage to 'groups'=

 RFH: The repos I build have no 'groups' defined, and we need to
 control livecdcreator/pungi and anaconda so that only the group (and
 its deps) is included.

 =Polish on the Anaconda / Installer=

  - With the F11 buildsys tools maybe we can now replace fedora-logos properly.

  - Anaconda on F11 seems to install from a bootable USB with less
 crashes -- will want to streamline the process.

 =Improve xs-config=

 There are several ways in which xs-config can be a lot less intrusive,
 for example using alternative init scripts (as we are doing with
 'pgsql-xs' vs 'postgresql').

 =FAQ=

 Why F11? Because the XS needs to be stable, and doesn't need bleeding
 edge, so targetting the LTS F11-RHEL6-CentOS path makes sense. The
 XS team would like to spend more time working on tools on top and
 less time tracking Fedora and regression-testing :-)

 Why are these packages out of sync with Fedora (missing F-12, etc)?
 The XS packages are still in flux and intrusive. And low-quality/risky
 from a Fedora PoV. As things stabilize in the future it might make
 sense to request inclusion.

I would encourage you to look at using F-12 rather than F-11. F-12
will be more likely to be the base for RHEL-6 than F-11 from what I've
heard so the move to CentOS 6 would be easier from that POV. F-12 is
pretty stable now, is feature complete and will be in beta in a week
or so and it will give you a 13 months of support from release rather
than the 6-7 that F-11 currently have rather than having something
based on a Fedora that's already EOL.

Either way let me know how i can help out.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] XS - rebase on F-11, notes and RHF

2009-10-08 Thread Peter Robinson
On Thu, Oct 8, 2009 at 12:26 PM, Martin Langhoff
martin.langh...@gmail.com wrote:
 On Thu, Oct 8, 2009 at 1:07 PM, Mathieu Bridon (bochecha)
 boche...@fedoraproject.org wrote:
 I think I told you that at the SugarCamp in Paris

 You weren't the only one :-) and in general it seemed to make sense.

 Ah, well. We're programmers, not mindreaders.

 After a bit of thought, initially at least, 0.7 will be based on F11.
 It is very hard for me to get my own code stable while the whole
 distro is in flux. Maybe there is a good opportunity for a F-12 rebase
 before 0.7.

Well if i can help out in any regards there feel free to ping me.

Cheers,
Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] [Sugar-devel] Notes on service discovery XS/XO

2009-04-22 Thread Peter Robinson
 I'm afraid that's not always true. Running Debian Lenny here without

 Paul, Iñaki, -- you guys are right.

 It also means that there's no point worrying about pushing extra
 records in the DNS responses.

It won't affect the XS side of things, but from the XO client side of
things the new gresolver interface in glib might be off interest to
people.

http://mysterion.org/~danw/blog/2009/04/gresolver

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel


Re: [Server-devel] Booting F9 kernels on XO...

2009-04-16 Thread Peter Robinson
 I have anaconda running on the XO, installs to usbkeys/booting from are
 fine with F9. F11's anaconda runs straight away, but opps when installing
 the rpms to the MMC card.

I think the issue your seeing with F11 should be fixed in the next
rawhide push (anaconda-11.5.0.45-1), I've been seeing the same issues.

Peter
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel