From a.travis at abdn.ac.uk Fri Sep 11 10:12:03 2009 From: a.travis at abdn.ac.uk (Tony Travis) Date: Fri, 11 Sep 2009 15:12:03 +0100 Subject: [Bio-linux-dev] 'biobuntu' Message-ID: <4AAA5AB3.6040607@abdn.ac.uk> Hello, Bio-Linux developers. I've updated my 'biobuntu' blueprint at: https://blueprints.launchpad.net/ubuntu/+spec/biobuntu Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt From a.travis at abdn.ac.uk Mon Sep 14 08:31:04 2009 From: a.travis at abdn.ac.uk (Tony Travis) Date: Mon, 14 Sep 2009 13:31:04 +0100 Subject: [Bio-linux-dev] [Fwd: Re: NuGO Desktop] Message-ID: <4AAE3788.7090801@abdn.ac.uk> Hello, Bio-Linux developers. Here's and update on our NBX-stick project... Have you had any feedback from Bio-Linux 5.0 users about problems with persistent sessions corrupting your ext2 "casper-rw" partition? It seems to work fine for 5-6 reboots, but then crashes badly :-( Using fsck shows "casper-rw" corruption after only one reboot from a persistent session. I'll test this by sacrificing my own NEBC USB-stick, and report back if I have the same problems. However, I got exactly the same problem on my 8GB Kingston DataTraveller, and I've low-level formatted the NBX-sticks using the Alcor_MP (Mass Production) tool to verify that there are no new (post manuf.) bad-blocks on the sticks but I still get corruption. Linux "badblocks -swp 1" didn't find any bad blocks on out NBX-stick, and neither did "mkfs -c ..." Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt -------------- next part -------------- An embedded message was scrubbed... From: Tony Travis Subject: Re: NuGO Desktop Date: Mon, 14 Sep 2009 13:20:12 +0100 Size: 2728 URL: -------------- next part -------------- An embedded message was scrubbed... From: "John Charly (BIGCAT)" Subject: NuGO Desktop Date: Mon, 14 Sep 2009 13:00:23 +0100 Size: 837854 URL: From tbooth at ceh.ac.uk Mon Sep 14 09:41:53 2009 From: tbooth at ceh.ac.uk (Tim Booth) Date: Mon, 14 Sep 2009 14:41:53 +0100 Subject: [Bio-linux-dev] [Fwd: Re: NuGO Desktop] In-Reply-To: <4AAE3788.7090801@abdn.ac.uk> References: <4AAE3788.7090801@abdn.ac.uk> Message-ID: <1252935713.13568.410.camel@barsukas> Hi Tony, Yes, we did run into this problem, I think. Stewart thought he'd solved it by modifying the shutdown part of the Casper scripts to force a sync. His notes say: --- Orginator: BT Date reported: 10 Dec 2008 Issue: USB stick problem on reboot seems like the disk failed to sync before it was ejected. Suggested fix if any: The script /etc/init.d/casper controls the shutdown and call on /sbin/casper-snapshot to sync disks. This is supposed to work for USB stick as well as DVD but it might need to be forced to sync. Fix implemented: Edited the /etc/init.d/casper file within the squash fs. In the subroutine do_stop created a condition that checks for the word "persistent" in /proc/cmdline which can be used to differentiate between usb and dvd. If system is running off usb the shutdown steps do a sync before allow ubuntu to complete the shutdown and NO MESSAGE is printed. With the DVD version it runs as before printing a message to remove the disk and close the tray before pressing enter. Tested and fixed. Fix implemented by: SH --- I remember at the time we reckoned the problem was that the live system printed a message saying "please remove the disk" before completing the shutdown. This is fine for read-only media but any read-write volumes were left still mounted and if you just pulled the stick out you left it corrupted. The solution was to ignore the message and finish shutting down with the stick still plugged in. Stewart's tweak makes this happen automatically. It's possible that Stewart missed something, but we've had few reports of problems since his fix. We tell people not to do important work on the live systems - it's so easy just to knock the thing and make it stop working and this leads to corruption of the FS. There is no automated fsck at boot, but so far manual fsck has been enough to get things working again in all problem cases I've seen. Cheers, TIM On Mon, 2009-09-14 at 13:31 +0100, Tony Travis wrote: > Hello, Bio-Linux developers. > > Here's and update on our NBX-stick project... > > Have you had any feedback from Bio-Linux 5.0 users about problems with > persistent sessions corrupting your ext2 "casper-rw" partition? > > It seems to work fine for 5-6 reboots, but then crashes badly :-( > > Using fsck shows "casper-rw" corruption after only one reboot from a > persistent session. I'll test this by sacrificing my own NEBC USB-stick, > and report back if I have the same problems. However, I got exactly the > same problem on my 8GB Kingston DataTraveller, and I've low-level > formatted the NBX-sticks using the Alcor_MP (Mass Production) tool to > verify that there are no new (post manuf.) bad-blocks on the sticks but > I still get corruption. Linux "badblocks -swp 1" didn't find any bad > blocks on out NBX-stick, and neither did "mkfs -c ..." > > Bye, > > Tony. > email message attachment > > -------- Forwarded Message -------- > > From: Tony Travis > > Reply-To: Travis, Tony > > To: John Charly (BIGCAT) > > Cc: nbx-dev at googlegroups.com > > Subject: Re: NuGO Desktop > > Date: Mon, 14 Sep 2009 13:20:12 +0100 > > > > John Charly (BIGCAT) wrote: > > > Hi Tony, I tried the the latest .iso for NuGO Linux from the NBX1. I > > > got a crash report for the package biolinux-artemis. Here I attached > > > the screenshot. > > > > Hello, Charly. > > > > OK, I know about that - It's a Bio-Linux problem that I'll have to fix. > > > > Thanks for trying out the iso! > > > > I'm still undecided about how best to format the USB sticks. The > > 'official' Ubuntu "usb-creator" has been back-ported to Hardy, and that > > is what I used to test NBX-sticks from the latest NuGO-Linux iso. > > > > The difference to the NEBC "bio-linux-usb-maker" is that "usb-creator" > > does NOT make a partition for an ext2 "casper-rw" filesytem. It uses a > > container file in the FAT32 partition to create an ext3 (journaling) > > filesystem. This seems a lot more robust, but there is a fix for > > "casper" that me and Philip de Groot tried out in Italy that might fix > > the problems with the ext2 "casper-rw" filesystem if we use 8.04.3 as > > the basis for remastering the Ubuntu CD instead of 8.04.1, which NEBC > > used: The 8.04.1 version is *known* to have problems with persistence! > > > > I'll keep you posted, but please try running the "usb-creator" on the > > NuGO-Linux DVD. You will need to create a symbolic link on the 'live' > > user's Desktop to do this: > > > > ln -s /dev/sdb1 cdrom.iso > > > > Then, specify this as the iso 'image' because "usb-creator" doesn't > > detect that it's running from the DVD for some reason. You run the > > official Ubuntu "usb-creator" from: > > > > System/Administration/Create a USB startup disk > > > > Any feedback would be very useful! > > > > Thanks, > > > > Tony. > > -- > > Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition > > and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK > > tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk > > mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt > > > email message attachment > > -------- Forwarded Message -------- > > From: John Charly (BIGCAT) > > To: Travis, Tony > > Subject: NuGO Desktop > > Date: Mon, 14 Sep 2009 13:00:23 +0100 > > > > Hi Tony, > > I tried the the latest .iso for NuGO Linux from the NBX1. I got a crash report for the package biolinux-artemis. Here I attached the screenshot. > > > > Thanks, > > > > Charly -- Tim Booth NERC Environmental Bioinformatics Centre at CEH Oxford +44 1865 281 975 From a.travis at abdn.ac.uk Mon Sep 14 11:31:53 2009 From: a.travis at abdn.ac.uk (Tony Travis) Date: Mon, 14 Sep 2009 16:31:53 +0100 Subject: [Bio-linux-dev] [Fwd: Re: NuGO Desktop] In-Reply-To: <1252935713.13568.410.camel@barsukas> References: <4AAE3788.7090801@abdn.ac.uk> <1252935713.13568.410.camel@barsukas> Message-ID: <4AAE61E9.2020000@abdn.ac.uk> Tim Booth wrote: > Hi Tony, > > Yes, we did run into this problem, I think. Stewart thought he'd solved > it by modifying the shutdown part of the Casper scripts to force a sync. > His notes say: Hello, Tim. I've seen the changes Stewart made but, in our tests, his code was never executed unless we changed it to unconditionally sync. His parsing of the kernel command line did not seem to work for us. However, the ext2 filesystem corruption still occured even when we forced casper to sync. It seems this problem is well-known in Ubuntu 8.04.1, and a patch was made to casper to deal with it by remounting the / and /cow filesystems read-only before shutting down: http://launchpadlibrarian.net/28114350/casper In addition, the official Ubuntu "usb-creator" backport uses a container file to hold an ext3 filesystem instead of mounting an ext2 partition: https://launchpad.net/ubuntu/hardy/i386/usb-creator/0.1.10~hardy1 I've watched the boot process carefully without quiet or splash on the kernel command-line, and ext3 completes any unfinished transactions at the next boot. This version seems quite robust in my experiments so far, but it takes a long time to create USB sticks using "usb-creator", so I've been modifying "bio-linux-usb-maker" to use a container file. I'll do some more experiments before making a decision about using an ext2 partition or not, but I'd welcome any feedback you can give about why the 'official' Ubuntu "usb-creator" does not use an ext2 partition? Part of my solution is to base our NuGO live DVD on Ubuntu 8.04.3, which has an updated kernel and initrd.gz on it. The casper problems seem to have been reported mainly in relation to the 8.04.1 live CD. We plan to encourage NuGO people to make use of persistent sessions on our NBX-sticks and also for them to use persistence to install updates, so our "casper-rw" filesystems have got to work reliably :-) Bye, Tony. > --- > Orginator: BT > Date reported: 10 Dec 2008 > Issue: USB stick problem on reboot seems like the disk failed to sync > before it was ejected. > > Suggested fix if any: The script /etc/init.d/casper controls the > shutdown and call on /sbin/casper-snapshot to sync disks. This is > supposed to work for USB stick as well as DVD but it might need to be > forced to sync. > Fix implemented: Edited the /etc/init.d/casper file within the squash > fs. In the subroutine do_stop created a condition that checks for the > word "persistent" in /proc/cmdline which can be used to differentiate > between usb and dvd. If system is running off usb the shutdown steps do > a sync before allow ubuntu to complete the shutdown and NO MESSAGE is > printed. With the DVD version it runs as before printing a message to > remove the disk and close the tray before pressing enter. Tested and > fixed. > Fix implemented by: SH > --- > > I remember at the time we reckoned the problem was that the live system > printed a message saying "please remove the disk" before completing the > shutdown. This is fine for read-only media but any read-write volumes > were left still mounted and if you just pulled the stick out you left it > corrupted. The solution was to ignore the message and finish shutting > down with the stick still plugged in. Stewart's tweak makes this happen > automatically. > > It's possible that Stewart missed something, but we've had few reports > of problems since his fix. We tell people not to do important work on > the live systems - it's so easy just to knock the thing and make it stop > working and this leads to corruption of the FS. There is no automated > fsck at boot, but so far manual fsck has been enough to get things > working again in all problem cases I've seen. > [...] -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt From tbooth at ceh.ac.uk Mon Sep 14 11:59:26 2009 From: tbooth at ceh.ac.uk (Tim Booth) Date: Mon, 14 Sep 2009 16:59:26 +0100 Subject: [Bio-linux-dev] [Fwd: Re: NuGO Desktop] In-Reply-To: <4AAE61E9.2020000@abdn.ac.uk> References: <4AAE3788.7090801@abdn.ac.uk> <1252935713.13568.410.camel@barsukas> <4AAE61E9.2020000@abdn.ac.uk> Message-ID: <1252943966.13568.427.camel@barsukas> Hi Tony, I hadn't appreciated that the issue was this complex. I'd avoided ext3 on the basis that journaling on flash media was liable to wear the thing out as well as slowing it down, but obviously the Ubuntu developers have reason to think otherwise. I'd still be wary of running any serious work from a 4-quid USB stick hanging out of the side of my machine, no matter how reliable you can make the software side of things. TIM From a.travis at abdn.ac.uk Tue Sep 15 05:36:12 2009 From: a.travis at abdn.ac.uk (Tony Travis) Date: Tue, 15 Sep 2009 10:36:12 +0100 Subject: [Bio-linux-dev] Taking data hostage Message-ID: <4AAF600C.3030208@abdn.ac.uk> Hello, As you know, I'm not a big fan of 'cloud' computing. I thought you might like to know that Google has now rejected the "data as hostage" business model used by commerical 'cloud' computing providers: http://arstechnica.com/tech-policy/news/2009/09/google-liberates-data-makes-it-easier-to-leave-the-cloud.ars There are several reasons why I'm opposed to buying computing resources from commerical 'cloud' (i.e. computer bureau) companies, and this is one of them. I think a more rational approach for NuGO/NTC is to use 'disruptive' p2p technology to aggregate the computing resources that we already own into a private 'cloud'... Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt From James.Kaput at fda.hhs.gov Tue Sep 15 08:55:18 2009 From: James.Kaput at fda.hhs.gov (Kaput, James) Date: Tue, 15 Sep 2009 08:55:18 -0400 Subject: [Bio-linux-dev] Taking data hostage In-Reply-To: <4AAF600C.3030208@abdn.ac.uk> References: <4AAF600C.3030208@abdn.ac.uk> Message-ID: <16C261088EE5734C945AB5658755A086044CB335@FMD3VS012.fda.gov> Tony Thanks for the alert The irony of course is that Google allows one to store photos, documents, and the like and is releasing Google WAVE (http://wave.google.com/) for collaborations (along with Google Health) So they are in fact arguing against something that they do - pretty slick :-) Jim -----Original Message----- From: Tony Travis [mailto:a.travis at abdn.ac.uk] Sent: Tuesday, September 15, 2009 4:36 AM To: nbx-dev+owner at googlegroups.com; bio-linux-dev at envgen.nox.ac.uk Cc: Kaput, James Subject: Taking data hostage Hello, As you know, I'm not a big fan of 'cloud' computing. I thought you might like to know that Google has now rejected the "data as hostage" business model used by commerical 'cloud' computing providers: http://arstechnica.com/tech-policy/news/2009/09/google-liberates-data-ma kes-it-easier-to-leave-the-cloud.ars There are several reasons why I'm opposed to buying computing resources from commerical 'cloud' (i.e. computer bureau) companies, and this is one of them. I think a more rational approach for NuGO/NTC is to use 'disruptive' p2p technology to aggregate the computing resources that we already own into a private 'cloud'... Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt From a.travis at abdn.ac.uk Wed Sep 23 13:00:21 2009 From: a.travis at abdn.ac.uk (Tony Travis) Date: Wed, 23 Sep 2009 18:00:21 +0100 Subject: [Bio-linux-dev] Rethink about using 64-bit Linux on NBX's Message-ID: <4ABA5425.1060504@abdn.ac.uk> Hello, NBX/Bio-Linux developers. I've been looking into the possibility of running a 32-bit Bio-Linux 'userland' under a 64-bit Linux kernel recently, because I want to run 64-bit software on our Beowulf servers and it would be useful to do the same on the NBX's. In the past, this required using a 32-bit 'chroot' environment, which was complicated and clumsy to set up. However, more recently the "ia32" libraries have been re-packaged for 64-bit Ubuntu: > i ia32-libs - ia32 shared libraries for use on amd64 and ia64 systems > v ia32-libs-gtk - > v ia32-libs-sdl - > i ia32-sun-java5-bin - Sun Java(TM) Runtime Environment (JRE) 5.0 (32-bit) > i ia32-sun-java6-bin - Sun Java(TM) Runtime Environment (JRE) 6 (32-bit) This makes it very easy to run existing Bio-Linux 32-bit binaries under a 64-bit kernel, but would also allow us to use the 64-bit version of R. Since we are at quite an early stage in upgrading the NBX's to Ubuntu 8.04 LTS, I wonder what you all think about upgrading them to 64-bit? I believe this might also of interest to other Bio-Linux users who want to use a 64-bit kernel, but don't want to wait for us to create source packages (or re-compile the sources in the existing NEBC 32-bit 'binary' packages). I've upgraded my own desktop PC, and one of our Beowulf servers to use 64-bit Ubuntu with the ia32 libraries and everything seems to work OK. The systems I tested had Bio-Linux packages installed (mainly) into /usr/local under 32-bit Ubuntu, then the system was upgraded to 64-bit Ubuntu overwriting the system partition but preserving /usr/local. The only problem I see at present is how to use the existing NEBC packages, which have control files specifying many of them as i386 architecture. Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt From mvaniersel at gmail.com Fri Sep 25 12:20:55 2009 From: mvaniersel at gmail.com (Martijn van Iersel) Date: Fri, 25 Sep 2009 18:20:55 +0200 Subject: [Bio-linux-dev] [nbx-dev] Rethink about using 64-bit Linux on NBX's In-Reply-To: <4ABA5425.1060504@abdn.ac.uk> References: <4ABA5425.1060504@abdn.ac.uk> Message-ID: <4ABCEDE7.4030906@gmail.com> Yes I agree, going 64 bit is the right way. But it seems to me that currently NuGO LDAP is blocking any kind of upgrade on the nbx'es. Until that problem is solved, this discussion is moot, really. At least that is my point of view. Martijn Tony Travis wrote: > Hello, NBX/Bio-Linux developers. > > I've been looking into the possibility of running a 32-bit Bio-Linux > 'userland' under a 64-bit Linux kernel recently, because I want to run > 64-bit software on our Beowulf servers and it would be useful to do the > same on the NBX's. In the past, this required using a 32-bit 'chroot' > environment, which was complicated and clumsy to set up. However, more > recently the "ia32" libraries have been re-packaged for 64-bit Ubuntu: > > >> i ia32-libs - ia32 shared libraries for use on amd64 and ia64 systems >> v ia32-libs-gtk - >> v ia32-libs-sdl - >> i ia32-sun-java5-bin - Sun Java(TM) Runtime Environment (JRE) 5.0 (32-bit) >> i ia32-sun-java6-bin - Sun Java(TM) Runtime Environment (JRE) 6 (32-bit) >> > > This makes it very easy to run existing Bio-Linux 32-bit binaries under > a 64-bit kernel, but would also allow us to use the 64-bit version of R. > > Since we are at quite an early stage in upgrading the NBX's to Ubuntu > 8.04 LTS, I wonder what you all think about upgrading them to 64-bit? > > I believe this might also of interest to other Bio-Linux users who want > to use a 64-bit kernel, but don't want to wait for us to create source > packages (or re-compile the sources in the existing NEBC 32-bit 'binary' > packages). > > I've upgraded my own desktop PC, and one of our Beowulf servers to use > 64-bit Ubuntu with the ia32 libraries and everything seems to work OK. > > The systems I tested had Bio-Linux packages installed (mainly) into > /usr/local under 32-bit Ubuntu, then the system was upgraded to 64-bit > Ubuntu overwriting the system partition but preserving /usr/local. The > only problem I see at present is how to use the existing NEBC packages, > which have control files specifying many of them as i386 architecture. > > Bye, > > Tony. > From a.travis at abdn.ac.uk Sun Sep 27 20:31:00 2009 From: a.travis at abdn.ac.uk (Tony Travis) Date: Mon, 28 Sep 2009 01:31:00 +0100 Subject: [Bio-linux-dev] [nbx-dev] Re: Rethink about using 64-bit Linux on NBX's In-Reply-To: <4ABCEDE7.4030906@gmail.com> References: <4ABA5425.1060504@abdn.ac.uk> <4ABCEDE7.4030906@gmail.com> Message-ID: <4AC003C4.7040608@abdn.ac.uk> Martijn van Iersel wrote: > Yes I agree, going 64 bit is the right way. > > But it seems to me that currently NuGO LDAP is blocking any kind of > upgrade on the nbx'es. Until that problem is solved, this discussion is > moot, really. At least that is my point of view. Hello, Martijn. Olivier Riche is taking over some of Ulrich's responsibilities, so that problem will soon be resolved. However, my question was do you think we should do something immediately about moving our upgraded NBX's to run a 64-bit kernel with a 32-bit Bio-Linux 5.0 userland? My reluctance to move to a 64-bit Ubuntu is because I want to make use of Bio-Linux 5.0, which is only available to us as 32-bit binary deb's. My attitude to running 64-bit Ubuntu has changed since I tried out the "ia32" libraries and I want to discuss the possibility of using 64-bit. The NBX hardware upgrade we're doing makes it very easy to clone an NBX, so I could send you a 512GB hard disk with the 64-bit kernel and 32-bit Bio-Linux 5.0 in /usr/local for you to try out on "nbx5". Please let me know what you think? Bye, Tony -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt From philip.degroot at wur.nl Mon Sep 28 03:41:21 2009 From: philip.degroot at wur.nl (Groot, Philip de) Date: Mon, 28 Sep 2009 09:41:21 +0200 Subject: [Bio-linux-dev] [nbx-dev] Re: Rethink about using 64-bit Linuxon NBX's References: <4ABA5425.1060504@abdn.ac.uk> <4ABCEDE7.4030906@gmail.com> <4AC003C4.7040608@abdn.ac.uk> Message-ID: Hello Tony, I am not readily available the coming weeks, so I am afraid that I cannot help a lot at the moment, but I do have a question: Why bother with ia32 libraries? If you go to the ubuntu download webpage, you can also select to download a 64bits Ubuntu version: http://www.ubuntu.com/GetUbuntu/download I wonder: how many packages will be missing if we install a 64 bits version? Source code is readily available, so we can try what compiles properly under 64 bits. Packages that cause problems need to be either fixed or executed in a 32bits environment (e.g. ia32). Ain't this approach better? I must honestly state that I don't exactly know how much work will be involved in this, but it must be given a try somewhere sometime anyway? Right? Regards, Dr. Philip de Groot Ph.D. Bioinformatics Researcher Wageningen University / TIFN Nutrigenomics Consortium Nutrition, Metabolism & Genomics Group Division of Human Nutrition PO Box 8129, 6700 EV Wageningen Visiting Address: Erfelijkheidsleer: De Valk, Building 304 Dreijenweg 2, 6703 HA Wageningen Room: 0052a T: +31-317-485786 F: +31-317-483342 E-mail: Philip.deGroot at wur.nl Internet: http://www.nutrigenomicsconsortium.nl http://humannutrition.wur.nl https://madmax.bioinformatics.nl ________________________________ From: bio-linux-dev-bounces at envgen.nerc-oxford.ac.uk on behalf of Tony Travis Sent: Mon 28-9-2009 2:31 To: nbx-dev at googlegroups.com Cc: bio-linux-dev at envgen.nox.ac.uk Subject: Re: [Bio-linux-dev] [nbx-dev] Re: Rethink about using 64-bit Linuxon NBX's Martijn van Iersel wrote: > Yes I agree, going 64 bit is the right way. > > But it seems to me that currently NuGO LDAP is blocking any kind of > upgrade on the nbx'es. Until that problem is solved, this discussion is > moot, really. At least that is my point of view. Hello, Martijn. Olivier Riche is taking over some of Ulrich's responsibilities, so that problem will soon be resolved. However, my question was do you think we should do something immediately about moving our upgraded NBX's to run a 64-bit kernel with a 32-bit Bio-Linux 5.0 userland? My reluctance to move to a 64-bit Ubuntu is because I want to make use of Bio-Linux 5.0, which is only available to us as 32-bit binary deb's. My attitude to running 64-bit Ubuntu has changed since I tried out the "ia32" libraries and I want to discuss the possibility of using 64-bit. The NBX hardware upgrade we're doing makes it very easy to clone an NBX, so I could send you a 512GB hard disk with the 64-bit kernel and 32-bit Bio-Linux 5.0 in /usr/local for you to try out on "nbx5". Please let me know what you think? Bye, Tony -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt _______________________________________________ Bio-Linux-dev mailing list Bio-Linux-dev at envgen.nox.ac.uk http://envgen.nox.ac.uk/mailman/listinfo/bio-linux-dev From a.travis at abdn.ac.uk Mon Sep 28 06:33:43 2009 From: a.travis at abdn.ac.uk (Travis, Tony) Date: Mon, 28 Sep 2009 11:33:43 +0100 Subject: [Bio-linux-dev] [nbx-dev] Re: Rethink about using 64-bit Linuxon NBX's In-Reply-To: References: <4ABA5425.1060504@abdn.ac.uk> <4ABCEDE7.4030906@gmail.com> <4AC003C4.7040608@abdn.ac.uk>, Message-ID: Hello, Philip. Perhaps I was not clear: I've installed 64-bit Ubuntu 8.04.3 LTS on one of our Beowulf servers at RINH, then installed the "ia32" libraries (and JRE's etc.) so that I can run existing 32-bit binaries under 64-bit Ubuntu. I'm not the only person using Bio-Linux 5.0 who would like to run a 64-bit Linux kernel, but does not want to wait until Bio-Linux is ported to 64-bit Ubuntu. My suggestion is a work-around so that I can migrate to 64-bit Ubunutu *now* instead of waiting for NEBC to port it to 64-bit (or port it myself). Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Rowett Institute of Nutrition and Health, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, Scotland, UK tel +44(0)1224 712751, fax +44(0)1224 716687, http://www.rowett.ac.uk mailto:a.travis at abdn.ac.uk, http://bioinformatics.rri.sari.ac.uk/~ajt ________________________________________ From: Groot, Philip de [philip.degroot at wur.nl] Sent: 28 September 2009 08:41 To: Travis, Tony; Bio-Linux technical discussion; nbx-dev at googlegroups.com Cc: bio-linux-dev at envgen.nox.ac.uk Subject: RE: [Bio-linux-dev] [nbx-dev] Re: Rethink about using 64-bit Linuxon NBX's Hello Tony, I am not readily available the coming weeks, so I am afraid that I cannot help a lot at the moment, but I do have a question: Why bother with ia32 libraries? If you go to the ubuntu download webpage, you can also select to download a 64bits Ubuntu version: http://www.ubuntu.com/GetUbuntu/download I wonder: how many packages will be missing if we install a 64 bits version? Source code is readily available, so we can try what compiles properly under 64 bits. Packages that cause problems need to be either fixed or executed in a 32bits environment (e.g. ia32). Ain't this approach better? I must honestly state that I don't exactly know how much work will be involved in this, but it must be given a try somewhere sometime anyway? Right? Regards, Dr. Philip de Groot Ph.D. Bioinformatics Researcher [...] The University of Aberdeen is a charity registered in Scotland, No SC013683.