Re: Donation of a Calxeda Highbank node
Generally speaking, our requirements for equipment to be used as buildd/porter machines are as follows: * reliability - The stable release manager requires that we operate three machines for each port: two buildd machines in different locations and one porter machine. These machines must be reliable. We already got ipa.debian.net, if David can give more nodes to Debian, then I think the requirement can be fulfilled. That we can do! Will I arrange this offline with someone in particular? * out of band management - We require the ability to manage the machines independently of their primary network interface: serial console or better, remotely-controllable power. David, could you also allow Debian admins into IPMI interface? Yup, shouldn¹t be a problem. * supportability - We require that the machines be commercialy available (within financial constraints) and that they be supportable through a warranty or post-warranty support or are otherwise easy to replace. Those machines are selling commercially now. * stability - We require that the machine's architecture have an actively-maintained stable kernel in the archive. Currently, there is an effort on going, as I explained above, but it is not yet finished, just starting: http://packages.debian.org/unstable/linux-image-3.9-1-armmp * environment - We require that packages critical for DSA operations be available: puppet, samhain, syslog-ng, ferm/pf, etc. The armhf port already contains those. We would prefer to house such equipment in one of the data centres where we have an existing presence (grnet (Greece), man-da (Germany), ubcece (Canada)) but we are amenable to a discussion regarding having a business host the equipment on our behalf as long as the above requirements are met. These will need to remotely accessed - we have a 24 node system that we use for customer to benchmark/test remotely. I can allocate one or two nodes on this but unfortunately couldn't provide a full chassis. In response to some of the other threads, I'll give some of the suggestions a try (kernel 3.9.1) and report back. Thanks, Dave Please let me know how you'd like to proceed. And thank you for your offer. It is *only* through the generous donation of time, equipment and/or funds from businesses and/or individuals such as yourself that volunteers are able to make the operating system known as Debian. Thanks again, Luca Debian System Administration Team Debian Hardware Donations Team -- Luca Filipozzi http://www.crowdrise.com/SupportDebian Regards, -- Héctor Orón -.. . -... .. .- -. -.. . ...- . .-.. --- .--. . .-. Debian ARM porter Debian ARM buildd admin DISCLAIMER: The information contained in this email is confidential and may be legally privileged. It is intended solely for the addressee and access by anyone else is unauthorised. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. When addressed to our customers, any information contained in this e-mail is subject to our standard terms and conditions. The views in this email may not necessarily reflect the views of the company. Please note that Boston uses the services of credit insurance agencies and cheque guarantee companies, therefore, customer details may be divulged as required by such companies. This information will be stored in their databases and may be exchanged with other credit reference agencies. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/cdf33e0d.60309%david.po...@boston.co.uk
Re: Donation of a Calxeda Highbank node
On Thu, Jun 27, 2013 at 09:29:12PM +, Luca Filipozzi wrote: We have to start somewhere, however. Having equipment on which to build, is a good start. I'll let the release team and the buildd team decide what to build, when and where. My goal is meet the supportability / reliability requirements that DSA has for an architecture. So, while you've brought up some excellent observations, let's not scare away the potential donor quite yet :) [/me waves at David] Well I did say that all you really need to do is get a working boot loader and kernel, and then the rest of wheezy should already work. How much simpler could it be? :) -- Len Sorensen -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130628143137.gv11...@csclub.uwaterloo.ca
Re: Donation of a Calxeda Highbank node
My main concern is that having a single node as buildd without another for development purposes means that we don't have easy means to keep testing for example kernel upgrades. Hey Guys, I stumbled upon this thread while trying to investigate if debian would work on the calxeda platform. We're one of calxedas partners and have specific customers who would really love to see debian working on ours/calxedas platform. Can I ask how far your efforts in porting have gone? Is there a wheezy release for highbank that we could test out internally? I would be more than happy to make some hardware available remotely for you to use as a second build/test server. We have genuine demand for debian on our platform. I'd also be happy to get involved in any way I can to help out. Thanks! Dave [please excuse the junk inserted by our mail server] DISCLAIMER: The information contained in this email is confidential and may be legally privileged. It is intended solely for the addressee and access by anyone else is unauthorised. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. When addressed to our customers, any information contained in this e-mail is subject to our standard terms and conditions. The views in this email may not necessarily reflect the views of the company. Please note that Boston uses the services of credit insurance agencies and cheque guarantee companies, therefore, customer details may be divulged as required by such companies. This information will be stored in their databases and may be exchanged with other credit reference agencies. -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/cdf0a78e.5fbe1%david.po...@boston.co.uk
Re: Donation of a Calxeda Highbank node
On Thu, Jun 27, 2013 at 07:27:54PM +, David Power wrote: Hey Guys, I stumbled upon this thread while trying to investigate if debian would work on the calxeda platform. We're one of calxedas partners and have specific customers who would really love to see debian working on ours/calxedas platform. Can I ask how far your efforts in porting have gone? Is there a wheezy release for highbank that we could test out internally? Well I highly doubt that will happen. Wheezy is released and uses 3.2 kernel which is certainly too old to work with this new hardware. Would http://packages.debian.org/jessie/linux-image-3.9-1-vexpress by any chance be a valid kernel image for the highbank? I see in the git logs of Linus's tree that there is a v7 multiplatform config that is supposed to cover the highbank, socfpga, mvebu, and vexpress in one kernel, so having a vexpress kernel in jessie could be a promising sign for the next major release. I would be more than happy to make some hardware available remotely for you to use as a second build/test server. We have genuine demand for debian on our platform. I'd also be happy to get involved in any way I can to help out. Well certainly I don't think there is a chance to be an officially supported piece of hardware until the next major release (in a couple of years, or whatever the release schedule is like these days). Now you could of course build a newer kernel, and install everything else from wheezy, and that should work fine, but it won't officially be Debian, and updates to the kernel for security and such would not be Debian's doing. Other than booting and the kernel, it is just yet another arm system though, so wheezy should work fine, otehr than the boot/kernel part. I am of course not in any way official Debian anything, just a long time user of Debian on many architectures and machine types, but I think I have a decent understanding of how things work. -- Len Sorensen -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130627211611.gu11...@csclub.uwaterloo.ca
Re: Donation of a Calxeda Highbank node
On Thu, Jun 27, 2013 at 07:27:54PM +, David Power wrote: On Tue, Apr 30, 2013 at 09:56:51AM +0300, Riku Voipio wrote: My main concern is that having a single node as buildd without another for development purposes means that we don't have easy means to keep testing for example kernel upgrades. Hey Guys, I stumbled upon this thread while trying to investigate if debian would work on the calxeda platform. We're one of calxedas partners and have specific customers who would really love to see debian working on ours/calxedas platform. Can I ask how far your efforts in porting have gone? Is there a wheezy release for highbank that we could test out internally? I would be more than happy to make some hardware available remotely for you to use as a second build/test server. We have genuine demand for debian on our platform. I'd also be happy to get involved in any way I can to help out. Hi Dave, We are more than happy to receive donations that assist us in properly addressing the buildd/porter requirements for the ARM architecture. Generally speaking, our requirements for equipment to be used as buildd/porter machines are as follows: * reliability - The stable release manager requires that we operate three machines for each port: two buildd machines in different locations and one porter machine. These machines must be reliable. * out of band management - We require the ability to manage the machines independently of their primary network interface: serial console or better, remotely-controllable power. * supportability - We require that the machines be commercialy available (within financial constraints) and that they be supportable through a warranty or post-warranty support or are otherwise easy to replace. * stability - We require that the machine's architecture have an actively-maintained stable kernel in the archive. * environment - We require that packages critical for DSA operations be available: puppet, samhain, syslog-ng, ferm/pf, etc. We would prefer to house such equipment in one of the data centres where we have an existing presence (grnet (Greece), man-da (Germany), ubcece (Canada)) but we are amenable to a discussion regarding having a business host the equipment on our behalf as long as the above requirements are met. Please let me know how you'd like to proceed. And thank you for your offer. It is *only* through the generous donation of time, equipment and/or funds from businesses and/or individuals such as yourself that volunteers are able to make the operating system known as Debian. Thanks again, Luca Debian System Administration Team Debian Hardware Donations Team -- Luca Filipozzi http://www.crowdrise.com/SupportDebian signature.asc Description: Digital signature
Re: Donation of a Calxeda Highbank node
Hello David, On Thu, 27 Jun 2013, Lennart Sorensen wrote: Would http://packages.debian.org/jessie/linux-image-3.9-1-vexpress by any chance be a valid kernel image for the highbank? I see in the git logs of Linus's tree that there is a v7 multiplatform config that is supposed to cover the highbank, socfpga, mvebu, and vexpress in one kernel, so having a vexpress kernel in jessie could be a promising sign for the next major release. Hector Oron sent us a kernel to test on the Hihgbank node we made available to Debian, but lack of knowledge on everything IPMI related means that we didn't dare to try it out yet: http://www.rtp-net.org/misc/linux-image-3.9-1-armmp_3.9-1~experimental.1_armhf.deb If you could try it out and give some feedback, it would be appreciated I'm sure. Cheers, -- Raphaël Hertzog ◈ Debian Developer Liberate the French translation of the Debian Administrator's Handbook: → http://www.ulule.com/liberation-cahier-admin-debian/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130627213430.ge12...@x230-buxy.home.ouaza.com
Re: Donation of a Calxeda Highbank node
On Thu, Jun 27, 2013 at 05:16:11PM -0400, Lennart Sorensen wrote: I am of course not in any way official Debian anything, just a long time user of Debian on many architectures and machine types, but I think I have a decent understanding of how things work. We have to start somewhere, however. Having equipment on which to build, is a good start. I'll let the release team and the buildd team decide what to build, when and where. My goal is meet the supportability / reliability requirements that DSA has for an architecture. So, while you've brought up some excellent observations, let's not scare away the potential donor quite yet :) [/me waves at David] Luca Debian System Administration Team Debian Hardware Donations Team -- Luca Filipozzi http://www.crowdrise.com/SupportDebian signature.asc Description: Digital signature
Re: Donation of a Calxeda Highbank node
Hello, 2013/6/27 Luca Filipozzi lfili...@debian.org: On Thu, Jun 27, 2013 at 07:27:54PM +, David Power wrote: On Tue, Apr 30, 2013 at 09:56:51AM +0300, Riku Voipio wrote: My main concern is that having a single node as buildd without another for development purposes means that we don't have easy means to keep testing for example kernel upgrades. Can I ask how far your efforts in porting have gone? Is there a wheezy release for highbank that we could test out internally? The armhf userland should run on those boxes, it already does in a chroot environment. ARM hardware enablement is missing in Debian, for example, custom IPMI Calxeda tools I think are not yet integrated in Debian, but Ubuntu have those available. There is also missing bootloader, and I do not think mainline u-boot/grub supports those boxes yet. Kernel wise, there is a current effort to migrate as much ARM platforms we have in Debian to armmp kernel (a.k.a. ARM multiplatform). Those kernels we have been trying to test on the donated node, but we are currently blocked on getting access to IPMI interface in case we need to recover from the disaster. If you can test armmp kernel and let us know something, that would really be a step forward. http://packages.debian.org/unstable/linux-image-3.9-1-armmp I would be more than happy to make some hardware available remotely for you to use as a second build/test server. We have genuine demand for debian on our platform. I'd also be happy to get involved in any way I can to help out. I think that is very welcome from ARM porters point of view. Thanks for the offer. Hi Dave, We are more than happy to receive donations that assist us in properly addressing the buildd/porter requirements for the ARM architecture. Generally speaking, our requirements for equipment to be used as buildd/porter machines are as follows: * reliability - The stable release manager requires that we operate three machines for each port: two buildd machines in different locations and one porter machine. These machines must be reliable. We already got ipa.debian.net, if David can give more nodes to Debian, then I think the requirement can be fulfilled. * out of band management - We require the ability to manage the machines independently of their primary network interface: serial console or better, remotely-controllable power. David, could you also allow Debian admins into IPMI interface? * supportability - We require that the machines be commercialy available (within financial constraints) and that they be supportable through a warranty or post-warranty support or are otherwise easy to replace. Those machines are selling commercially now. * stability - We require that the machine's architecture have an actively-maintained stable kernel in the archive. Currently, there is an effort on going, as I explained above, but it is not yet finished, just starting: http://packages.debian.org/unstable/linux-image-3.9-1-armmp * environment - We require that packages critical for DSA operations be available: puppet, samhain, syslog-ng, ferm/pf, etc. The armhf port already contains those. We would prefer to house such equipment in one of the data centres where we have an existing presence (grnet (Greece), man-da (Germany), ubcece (Canada)) but we are amenable to a discussion regarding having a business host the equipment on our behalf as long as the above requirements are met. Please let me know how you'd like to proceed. And thank you for your offer. It is *only* through the generous donation of time, equipment and/or funds from businesses and/or individuals such as yourself that volunteers are able to make the operating system known as Debian. Thanks again, Luca Debian System Administration Team Debian Hardware Donations Team -- Luca Filipozzi http://www.crowdrise.com/SupportDebian Regards, -- Héctor Orón -.. . -... .. .- -. -.. . ...- . .-.. --- .--. . .-. Debian ARM porter Debian ARM buildd admin -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CAODfWeH_YXZpH=vsjgkxteo4t-8nqhrked8r1xbse4nzcat...@mail.gmail.com
Re: Donation of a Calxeda Highbank node
On Thu, 2013-06-27 at 17:16 -0400, Lennart Sorensen wrote: On Thu, Jun 27, 2013 at 07:27:54PM +, David Power wrote: Hey Guys, I stumbled upon this thread while trying to investigate if debian would work on the calxeda platform. We're one of calxedas partners and have specific customers who would really love to see debian working on ours/calxedas platform. Can I ask how far your efforts in porting have gone? Is there a wheezy release for highbank that we could test out internally? Well I highly doubt that will happen. Wheezy is released and uses 3.2 kernel which is certainly too old to work with this new hardware. Would http://packages.debian.org/jessie/linux-image-3.9-1-vexpress by any chance be a valid kernel image for the highbank? I see in the git logs of Linus's tree that there is a v7 multiplatform config that is supposed to cover the highbank, socfpga, mvebu, and vexpress in one kernel, so having a vexpress kernel in jessie could be a promising sign for the next major release. [...] The 'armmp' flavour is the one with a multiplatform configuration. All the other armhf kernel flavours will be dropped once it's ready to replace them. Ben. -- Ben Hutchings Knowledge is power. France is bacon. signature.asc Description: This is a digitally signed message part
Re: Donation of a Calxeda Highbank node
On Sun, Apr 28, 2013 at 10:30:27AM -0700, Steve Langasek wrote: On Sat, Apr 27, 2013 at 09:32:59AM +0200, Peter Palfrader wrote: The impression I got during the brief from the arm porters is that it is so far unclear how well Debian will run on this nice shiney thing. So for now it's just a test box/early porting box, and the policies and procedures that come with DSAing a machine would be more a hindrance than an asset during that stage. That's fair, though I think the explicit goal should be to get it supported by Debian *so that* it can be used as a buildd. I agree. Given that highbank is well supported mainline it and debian being popular at server front, this platform should be a good fit for debian. Yes, understood; and I propose that buildd is the best use for it in the long term. I think so too, using server hardware for buildd's as opposed to developer boards for mobile usage sounds like a major win in the robustness front. My main concern is that having a single node as buildd without another for development purposes means that we don't have easy means to keep testing for example kernel upgrades. Riku -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130430065651.ga29...@afflict.kos.to
Re: Donation of a Calxeda Highbank node
Hello, 2013/4/26 Raphael Hertzog hert...@debian.org: On Fri, 26 Apr 2013, Hector Oron wrote: * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. * Buildd hat: The machine shall not be used to run buildd software to build official packages. Did you get those answers from the relevant teams? Was there a rationale for those answers? Yes, I did check with DSA. With porters and buildd team, I did not check, besides saying here, so whoever is interested is able to forward their opinion. For the rationale, you probably already have enough information on this discussion thread. Now, Raphael, apparently you got a contact already with the donator and a root account on the server machine. Would you be willing to admin or handout root access to few ARM porters, so DD accounts could either be exported or create local accounts for people willing to play with the machine? I can give root rights to porters who want to administrate the machine, yes. I'd rather not have to take care of routine maintenance but I can stay as a backup in case of need. To be a useful porter machine, it would be nice if all DD can have access to it. But if it's not administrated by DSA, I don't know if the newly announced self-served chroot service can be setup on this machine. Yes, we can do that. I'll contact you privately and I'll take over the setup. If any porter fancies to admin such machine, please contact Raphael and myself. Regards -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CAODfWeEauh5r+Kn6EEYm=JiC5qzT4diodqcuzfkpAEPiJe=v...@mail.gmail.com
Re: Donation of a Calxeda Highbank node
On Sat, Apr 27, 2013 at 09:32:59AM +0200, Peter Palfrader wrote: On Fri, 26 Apr 2013, Steve Langasek wrote: On Fri, Apr 26, 2013 at 02:49:05PM +0200, Hector Oron wrote: Thanks, that is a very kind offer and with ARM hat on, we cannot reject the offer, it makes it very interesting as a playground machine. However, let me make some points here: * ARM porters hat: It is very interesting machine, and very useful to start experimenting with it as Debian is seeking for a full Calxeda chassis. * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. Why in the world not? I'm sure there's no requirement for debian.org machines to be hardware owned by Debian. The s390 porter machines/buildds certainly aren't; I don't see why this machine would necessarily *not* be a d.o machine managed by DSA. Of course if it's going to be DSA-administered, I'm sure DSA would want exclusive admin rights on the machine; but that's just common sense, and AIUI not excluded by the offer. The impression I got during the brief from the arm porters is that it is so far unclear how well Debian will run on this nice shiney thing. So for now it's just a test box/early porting box, and the policies and procedures that come with DSAing a machine would be more a hindrance than an asset during that stage. That's fair, though I think the explicit goal should be to get it supported by Debian *so that* it can be used as a buildd. FWIW, Ubuntu 13.04 ships with support for Highbank using (AIUI) an unmodified upstream kernel. So while there may be some porting work to be done before Debian runs out of the box on it, there's prior art and I don't imagine the porting work required will be huge. Also, if it were a d.o system, it would be /either/ a porterbox /or/ a buildd, not both. Yes, understood; and I propose that buildd is the best use for it in the long term. Whereas, as long as it's a test/play system run by the porters presumably, they can stress test is at needed, maybe run a (non-official?) buildd, while also providing porter chroots. Once we have Debian running properly on this kind of HW, I wouldn't mind taking over the machine. Though, to be really useful, we probably will try to get more than one instance, one for a porterbox, and two - ideally in different locations - for autobuilding packages. That would obviously be a pretty awesome end state. But it also obviously depends on the generosity of other donors. In the meantime, I hope we don't let the perfect be the enemy of the good here, because having even just one of these nodes available is already VERY good for us - provided we don't let it go to waste. Cheers, -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developerhttp://www.debian.org/ slanga...@ubuntu.com vor...@debian.org signature.asc Description: Digital signature
Re: Donation of a Calxeda Highbank node
On Sun, 2013-04-28 at 10:30 -0700, Steve Langasek wrote: On Sat, Apr 27, 2013 at 09:32:59AM +0200, Peter Palfrader wrote: On Fri, 26 Apr 2013, Steve Langasek wrote: On Fri, Apr 26, 2013 at 02:49:05PM +0200, Hector Oron wrote: Thanks, that is a very kind offer and with ARM hat on, we cannot reject the offer, it makes it very interesting as a playground machine. However, let me make some points here: * ARM porters hat: It is very interesting machine, and very useful to start experimenting with it as Debian is seeking for a full Calxeda chassis. * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. Why in the world not? I'm sure there's no requirement for debian.org machines to be hardware owned by Debian. The s390 porter machines/buildds certainly aren't; I don't see why this machine would necessarily *not* be a d.o machine managed by DSA. Of course if it's going to be DSA-administered, I'm sure DSA would want exclusive admin rights on the machine; but that's just common sense, and AIUI not excluded by the offer. The impression I got during the brief from the arm porters is that it is so far unclear how well Debian will run on this nice shiney thing. So for now it's just a test box/early porting box, and the policies and procedures that come with DSAing a machine would be more a hindrance than an asset during that stage. That's fair, though I think the explicit goal should be to get it supported by Debian *so that* it can be used as a buildd. FWIW, Ubuntu 13.04 ships with support for Highbank using (AIUI) an unmodified upstream kernel. So while there may be some porting work to be done before Debian runs out of the box on it, there's prior art and I don't imagine the porting work required will be huge. [...] The plan of record for Debian armhf kernel support is to introduce a multi-platform 'armmp' flavour starting with Linux 3.9. So far as I know, it would be fairly easy to include Highbank support in that. Boot loader support is presumably going to require more work. Ben. -- Ben Hutchings Sturgeon's Law: Ninety percent of everything is crap. signature.asc Description: This is a digitally signed message part
Re: Donation of a Calxeda Highbank node
On Fri, 26 Apr 2013, Steve Langasek wrote: On Fri, Apr 26, 2013 at 02:49:05PM +0200, Hector Oron wrote: Thanks, that is a very kind offer and with ARM hat on, we cannot reject the offer, it makes it very interesting as a playground machine. However, let me make some points here: * ARM porters hat: It is very interesting machine, and very useful to start experimenting with it as Debian is seeking for a full Calxeda chassis. * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. Why in the world not? I'm sure there's no requirement for debian.org machines to be hardware owned by Debian. The s390 porter machines/buildds certainly aren't; I don't see why this machine would necessarily *not* be a d.o machine managed by DSA. Of course if it's going to be DSA-administered, I'm sure DSA would want exclusive admin rights on the machine; but that's just common sense, and AIUI not excluded by the offer. The impression I got during the brief from the arm porters is that it is so far unclear how well Debian will run on this nice shiney thing. So for now it's just a test box/early porting box, and the policies and procedures that come with DSAing a machine would be more a hindrance than an asset during that stage. Also, if it were a d.o system, it would be /either/ a porterbox /or/ a buildd, not both. Whereas, as long as it's a test/play system run by the porters presumably, they can stress test is at needed, maybe run a (non-official?) buildd, while also providing porter chroots. Once we have Debian running properly on this kind of HW, I wouldn't mind taking over the machine. Though, to be really useful, we probably will try to get more than one instance, one for a porterbox, and two - ideally in different locations - for autobuilding packages. Cheers, weasel -- | .''`. ** Debian ** Peter Palfrader | : :' : The universal http://www.palfrader.org/ | `. `' Operating System | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130427073259.gx23...@anguilla.noreply.org
Re: Donation of a Calxeda Highbank node
On Fri, 26 Apr 2013, Raphael Hertzog wrote: To be a useful porter machine, it would be nice if all DD can have access to it. But if it's not administrated by DSA, I don't know if the newly announced self-served chroot service can be setup on this machine. Neither of these require the machine to be DSA administered. We can help in setting these up initially, and I assume Hector would be happy to take a more hands-on approach there. Cheers -- | .''`. ** Debian ** Peter Palfrader | : :' : The universal http://www.palfrader.org/ | `. `' Operating System | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130427073625.gy23...@anguilla.noreply.org
Re: Donation of a Calxeda Highbank node
Hello, 2013/4/12 Raphael Hertzog hert...@debian.org: OffensiveSecurity owns a Calxeda Highbank cluster of ARM machines[1] and is willing to dedicate one of the nodes to Debian. Thanks, that is a very kind offer and with ARM hat on, we cannot reject the offer, it makes it very interesting as a playground machine. However, let me make some points here: * ARM porters hat: It is very interesting machine, and very useful to start experimenting with it as Debian is seeking for a full Calxeda chassis. * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. * Buildd hat: The machine shall not be used to run buildd software to build official packages. * Kernel hat: As Ben already stated, ARM multiplatform kernel should already support this machine. Now, Raphael, apparently you got a contact already with the donator and a root account on the server machine. Would you be willing to admin or handout root access to few ARM porters, so DD accounts could either be exported or create local accounts for people willing to play with the machine? Best regards -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CAODfWeG=O4VLr5HSOMS78J+aF-7sr0gOx8bqWe=qutqes-y...@mail.gmail.com
Re: Donation of a Calxeda Highbank node
Hector Oron hector.o...@gmail.com writes: Hello, 2013/4/12 Raphael Hertzog hert...@debian.org: OffensiveSecurity owns a Calxeda Highbank cluster of ARM machines[1] and is willing to dedicate one of the nodes to Debian. Thanks, that is a very kind offer and with ARM hat on, we cannot reject the offer, it makes it very interesting as a playground machine. However, let me make some points here: * ARM porters hat: It is very interesting machine, and very useful to start experimenting with it as Debian is seeking for a full Calxeda chassis. * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. * Buildd hat: The machine shall not be used to run buildd software to build official packages. * Kernel hat: As Ben already stated, ARM multiplatform kernel should already support this machine. yes, it should. My patched 3.8 multiplatform is more or less booting with qemu emulating Highbank machine but I'm getting either some warnings or it just hangs, don't know why. So, if we're lucky, it's something related to qemu and it'll just work, if we're not some patching will be needed. Now, Raphael, apparently you got a contact already with the donator and a root account on the server machine. Would you be willing to admin or handout root access to few ARM porters, so DD accounts could either be exported or create local accounts for people willing to play with the machine? It would also be interesting to know if it's possible to test a kernel to make sure that the multiplatform kernel will boot/work on it. Thanks, Arnaud -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/87ehdx2z71@lebrac.rtp-net.org
Re: Donation of a Calxeda Highbank node
On Fri, 26 Apr 2013, Hector Oron wrote: * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. * Buildd hat: The machine shall not be used to run buildd software to build official packages. Did you get those answers from the relevant teams? Was there a rationale for those answers? Now, Raphael, apparently you got a contact already with the donator and a root account on the server machine. Would you be willing to admin or handout root access to few ARM porters, so DD accounts could either be exported or create local accounts for people willing to play with the machine? I can give root rights to porters who want to administrate the machine, yes. I'd rather not have to take care of routine maintenance but I can stay as a backup in case of need. To be a useful porter machine, it would be nice if all DD can have access to it. But if it's not administrated by DSA, I don't know if the newly announced self-served chroot service can be setup on this machine. Cheers, -- Raphaël Hertzog ◈ Debian Developer Get the Debian Administrator's Handbook: → http://debian-handbook.info/get/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130426140403.ga4...@x230-buxy.home.ouaza.com
Re: Donation of a Calxeda Highbank node
On Fri, Apr 26, 2013 at 02:49:05PM +0200, Hector Oron wrote: 2013/4/12 Raphael Hertzog hert...@debian.org: OffensiveSecurity owns a Calxeda Highbank cluster of ARM machines[1] and is willing to dedicate one of the nodes to Debian. Thanks, that is a very kind offer and with ARM hat on, we cannot reject the offer, it makes it very interesting as a playground machine. However, let me make some points here: * ARM porters hat: It is very interesting machine, and very useful to start experimenting with it as Debian is seeking for a full Calxeda chassis. * DSA hat: The machine shall not be a debian.org machine, so DSA could export accounts if requested. Why in the world not? I'm sure there's no requirement for debian.org machines to be hardware owned by Debian. The s390 porter machines/buildds certainly aren't; I don't see why this machine would necessarily *not* be a d.o machine managed by DSA. Of course if it's going to be DSA-administered, I'm sure DSA would want exclusive admin rights on the machine; but that's just common sense, and AIUI not excluded by the offer. * Buildd hat: The machine shall not be used to run buildd software to build official packages. Again, why not? As far as I can see, this would be the single best use that a Highbank node could be put to by the project. As http://release.debian.org/wheezy/arch_qualify.html shows, we have quite a few more buildds deployed for both armel and armhf than is optimal - and in the case of armhf, we have so many that we're technically exceeding the release requirements. A single Highbank node could replace at least 2 of the fastest buildds we currently have in production... and probably somewhere between 4 and 8 of the slow ones. It absolutely makes sense to leverage this donation as a buildd and decommision/reallocate some of the lower-powered buildds, increasing the reliability of the buildd pool and reducing the total machine management overhead. Being the only highbank machine we have access to does imply some logistical challenges for ensuring that we continue to have good kernel support during the development cycle, so that release+1 will be supportable on the box. But this is a problem we've dealt with before, and certainly in this case the upstream kernel support is quite good, which I think makes this a fairly low risk. Cheers, -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developerhttp://www.debian.org/ slanga...@ubuntu.com vor...@debian.org signature.asc Description: Digital signature
Donation of a Calxeda Highbank node
Hello, OffensiveSecurity owns a Calxeda Highbank cluster of ARM machines[1] and is willing to dedicate one of the nodes to Debian. I believe it would make a relatively fast ARM autobuilder or porter box. The specs of the hardware (4GB RAM, 500 GB SATA drives, 4 Cortex A9 cores at 1.1 to 1.4 GHz) are better than most (all?) of Debian's ARM boxes. However this host runs on Ubuntu by default (12.10 currently). AFAIK Debian's kernel doesn't support such machines yet, Ubuntu has a special ARM flavor called highbank [2]. So I'm not sure how we could handle this donation right now (putting debian-admin@ in the loop to see what requirements they have at this level, and debian-kernel@ to see whether such a flavor would be doable for Debian too). I am however running Debian armel/armhf chroots on such a machine without any problem. So it seems that the kernel is the only problematic part. I do have root access to the system and I can pass it to DSA or any porter who would need access to the host. The thing supports IPMI and various management tools but they are not currently setup and I'm not familiar with them. If needed, with guidance, and if it supports proper per-node delegation, we could imagine giving access to DSA to this management tool. If you have questions, feel free to ask. I'll do my best to answer. Some more infos: # cat /proc/cpuinfo Processor : ARMv7 Processor rev 0 (v7l) processor : 0 BogoMIPS: 2190.54 processor : 1 BogoMIPS: 2190.54 processor : 2 BogoMIPS: 2190.54 processor : 3 BogoMIPS: 2190.54 Features: swp half thumb fastmult vfp edsp neon vfpv3 tls CPU implementer : 0x41 CPU architecture: 7 CPU variant : 0x3 CPU part: 0xc09 CPU revision: 0 Hardware: Highbank Revision: Serial : # df -h Filesystem Size Used Avail Use% Mounted on /dev/sda2 455G 2.1G 430G 1% / udev2.0G 4.0K 2.0G 1% /dev tmpfs 809M 176K 808M 1% /run none5.0M 0 5.0M 0% /run/lock none2.0G 0 2.0G 0% /run/shm none100M 0 100M 0% /run/user /dev/sda189M 53M 32M 63% /boot # free -m total used free sharedbuffers cached Mem: 4040 1695 2344 0266 1245 -/+ buffers/cache:183 3856 Swap: 3944 0 3944 # uname -a Linux arm08 3.5.0-22-highbank #33-Ubuntu SMP Thu Jan 3 01:05:04 UTC 2013 armv7l armv7l armv7l GNU/Linux # cat /etc/os-release NAME=Ubuntu VERSION=12.10, Quantal Quetzal ID=ubuntu ID_LIKE=debian PRETTY_NAME=Ubuntu quantal (12.10) VERSION_ID=12.10 Cheers, [1] More precisely a SystemFabriCore: http://www.systemfabricworks.com/products/systemfabricore [2] http://ncommander.blogspot.fr/2012/06/announcement-of-calxeda-highbank-images.html https://wiki.ubuntu.com/ARM/Server/Install -- Raphaël Hertzog ◈ Debian Developer Get the Debian Administrator's Handbook: → http://debian-handbook.info/get/ -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130412151619.ga21...@x230-buxy.home.ouaza.com
Re: Donation of a Calxeda Highbank node
On Fri, Apr 12, 2013 at 05:16:19PM +0200, Raphael Hertzog wrote: Hello, OffensiveSecurity owns a Calxeda Highbank cluster of ARM machines[1] and is willing to dedicate one of the nodes to Debian. I believe it would make a relatively fast ARM autobuilder or porter box. The specs of the hardware (4GB RAM, 500 GB SATA drives, 4 Cortex A9 cores at 1.1 to 1.4 GHz) are better than most (all?) of Debian's ARM boxes. However this host runs on Ubuntu by default (12.10 currently). AFAIK Debian's kernel doesn't support such machines yet, Ubuntu has a special ARM flavor called highbank [2]. So I'm not sure how we could handle this donation right now (putting debian-admin@ in the loop to see what requirements they have at this level, and debian-kernel@ to see whether such a flavor would be doable for Debian too). [...] Starting with Linux 3.9 there will be an 'armmp' (multi-platform) flavour for armhf. All support for new ARMv7 platforms should be added to that flavour by enabling the relevant drivers and DeviceTree files, not by adding new flavours. Highbank seems to be at least mostly ready for multi-platform. Ben. -- Ben Hutchings We get into the habit of living before acquiring the habit of thinking. - Albert Camus -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130412181034.go2...@decadent.org.uk