From aleleb at sahfos.ac.uk Fri Jul 5 09:29:11 2013 From: aleleb at sahfos.ac.uk (Alec Colebrook-Clark) Date: Fri, 5 Jul 2013 14:29:11 +0100 Subject: [Bio-Linux] Bandwidth Issue Message-ID: Hi We have an issue at the moment where Biolinux will consume every bit of available bandwidth when it is downloading. This has huge impacts on the rest of the network. Is anyone else experiencing this? And does anyone have any idea how to cope/fix the issue? Thanks Alec -------------- next part -------------- An HTML attachment was scrubbed... URL: From tony.travis at ed.ac.uk Fri Jul 5 09:59:28 2013 From: tony.travis at ed.ac.uk (Tony Travis) Date: Fri, 05 Jul 2013 14:59:28 +0100 Subject: [Bio-Linux] Bandwidth Issue In-Reply-To: References: Message-ID: <51D6D140.2040008@ed.ac.uk> On 05/07/13 14:29, Alec Colebrook-Clark wrote: > Hi > > > > We have an issue at the moment where Biolinux will consume every bit of > available bandwidth when it is downloading. This has huge impacts on the > rest of the network. Is anyone else experiencing this? And does anyone > have any idea how to cope/fix the issue? Hi, Alec. Not quite sure what you mean "when it is downloading": Do you mean when people are downloading the Bio-Linux .iso file, or when Bio-Linux is running programs that download files? Tony. From tbooth at ceh.ac.uk Fri Jul 5 10:01:27 2013 From: tbooth at ceh.ac.uk (Tim Booth) Date: Fri, 5 Jul 2013 15:01:27 +0100 Subject: [Bio-Linux] Bandwidth Issue In-Reply-To: References: Message-ID: <1373032887.17766.335.camel@balisaur> Hi Alec, Is this because you have many BL boxes doing updates at once and hammering the WAN, or do you want to just slow down one individual box? In the former case, you probably want to deploy the apt-cacher-ng package on a server machine to ensure that packages are only fetched over the internet once and then just copied over the LAN from the cache. In the latter case, search the manpage for apt.conf for "Dl-Limit" and "Pipeline-Depth" which allow you to throttle bandwidth usage on any machine. Let me know if you need any more pointers on the above. Cheers, TIM On Fri, 2013-07-05 at 14:29 +0100, Alec Colebrook-Clark wrote: > Hi > > > > We have an issue at the moment where Biolinux will consume every bit > of available bandwidth when it is downloading. This has huge impacts > on the rest of the network. Is anyone else experiencing this? And does > anyone have any idea how to cope/fix the issue? > > > > Thanks > > Alec > > > > -- Tim Booth NERC Environmental Bioinformatics Centre Centre for Ecology and Hydrology Maclean Bldg, Benson Lane Crowmarsh Gifford Wallingford, England OX10 8BB http://nebc.nerc.ac.uk +44 1491 69 2705 From aleleb at sahfos.ac.uk Fri Jul 12 08:42:00 2013 From: aleleb at sahfos.ac.uk (Alec Colebrook-Clark) Date: Fri, 12 Jul 2013 13:42:00 +0100 Subject: [Bio-Linux] Bandwidth cont. Message-ID: Hi, Sorry for the delay, don't seem to get these as often as I like, have played with the settings and should have rectified that! @Tony - By download I mean the actual biolinux machine is downloading a file onto a location on the local machine. @TIM - the problem is a single box hammering the WAN. I tried to throttle the machine but it only effected the its LAN capabilities and not the download (which is happening over the same connection). Is there any way I can trace what is happening? Thanks for the help Alec -------------- next part -------------- An HTML attachment was scrubbed... URL: From tony.travis at abdn.ac.uk Fri Jul 12 08:53:07 2013 From: tony.travis at abdn.ac.uk (Tony Travis) Date: Fri, 12 Jul 2013 13:53:07 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: References: Message-ID: <51DFFC33.90704@abdn.ac.uk> On 12/07/13 13:42, Alec Colebrook-Clark wrote: > Hi, > > > > Sorry for the delay, don?t seem to get these as often as I like, have > played with the settings and should have rectified that! > > > > @Tony - By download I mean the actual biolinux machine is downloading a > file onto a location on the local machine. Hi, Alec. You can limit the bandwidth used by "wget" > --limit-rate=amount > Limit the download speed to amount bytes per second. Amount may be expressed in bytes, kilobytes with the k suffix, > or megabytes with the m suffix. For example, --limit-rate=20k will limit the retrieval rate to 20KB/s. This is > useful when, for whatever reason, you don't want Wget to consume the entire available bandwidth. Similar with "sftp": > -l limit > Limits the used bandwidth, specified in Kbit/s. How are your doing the downloads? > @TIM ? the problem is a single box hammering the WAN. I tried to > throttle the machine but it only effected the its LAN capabilities and > not the download (which is happening over the same connection). Is there > any way I can trace what is happening? Use "netstat"/"iftop" to monitor the connections/network traffic? HTH, Tony. -- Dr. A.J.Travis, University of Aberdeen, Institute of Biological and Environmental Sciences, Cruickshank Building, St. Machar Drive, Aberdeen AB24 3UU, Scotland, UK. tel +44(0)1224 272700, fax +44 (0)1224 272 396 http://www.abdn.ac.uk, mailto:tony.travis at abdn.ac.uk, skype:ajtravis From aleleb at sahfos.ac.uk Fri Jul 12 08:59:39 2013 From: aleleb at sahfos.ac.uk (Alec Colebrook-Clark) Date: Fri, 12 Jul 2013 13:59:39 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: <51DFFC33.90704@abdn.ac.uk> References: <51DFFC33.90704@abdn.ac.uk> Message-ID: Hi Tony, Ive been doing the downloads by sshing into the machine and using sftp or wget. I have already come across these limit options with little success but I shall investigate more. And thanks for the heads up on netstat. I have tried iftop but it just confirms its using all bandwidth (unless im using the tool incorrectly.) Thanks Alec -----Original Message----- From: bio-linux-bounces at nebclists.nerc.ac.uk [mailto:bio-linux-bounces at nebclists.nerc.ac.uk] On Behalf Of Tony Travis Sent: 12 July 2013 13:53 To: bio-linux at nebclists.nerc.ac.uk Subject: Re: [Bio-Linux] Bandwidth cont. On 12/07/13 13:42, Alec Colebrook-Clark wrote: > Hi, > > > > Sorry for the delay, don't seem to get these as often as I like, have > played with the settings and should have rectified that! > > > > @Tony - By download I mean the actual biolinux machine is downloading > a file onto a location on the local machine. Hi, Alec. You can limit the bandwidth used by "wget" > --limit-rate=amount > Limit the download speed to amount bytes per second. Amount may be expressed in bytes, kilobytes with the k suffix, > or megabytes with the m suffix. For example, --limit-rate=20k will limit the retrieval rate to 20KB/s. This is > useful when, for whatever reason, you don't want Wget to consume the entire available bandwidth. Similar with "sftp": > -l limit > Limits the used bandwidth, specified in Kbit/s. How are your doing the downloads? > @TIM - the problem is a single box hammering the WAN. I tried to > throttle the machine but it only effected the its LAN capabilities and > not the download (which is happening over the same connection). Is > there any way I can trace what is happening? Use "netstat"/"iftop" to monitor the connections/network traffic? HTH, Tony. -- Dr. A.J.Travis, University of Aberdeen, Institute of Biological and Environmental Sciences, Cruickshank Building, St. Machar Drive, Aberdeen AB24 3UU, Scotland, UK. tel +44(0)1224 272700, fax +44 (0)1224 272 396 http://www.abdn.ac.uk, mailto:tony.travis at abdn.ac.uk, skype:ajtravis _______________________________________________ Bio-Linux mailing list Bio-Linux at nebclists.nerc.ac.uk http://nebclists.nerc.ac.uk/mailman/listinfo/bio-linux From tony.travis at abdn.ac.uk Fri Jul 12 09:07:03 2013 From: tony.travis at abdn.ac.uk (Tony Travis) Date: Fri, 12 Jul 2013 14:07:03 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: References: <51DFFC33.90704@abdn.ac.uk> Message-ID: <51DFFF77.2090201@abdn.ac.uk> On 12/07/13 13:59, Alec Colebrook-Clark wrote: > Hi Tony, > > Ive been doing the downloads by sshing into the machine and using > sftp or wget. I have already come across these limit options with > little success but I shall investigate more. And thanks for the heads > up on netstat. I have tried iftop but it just confirms its using all > bandwidth (unless im using the tool incorrectly.) Hi, Alec. Have you tried nice'ing the download program? nice -n 19 sftp ... To throttle at the hardware level, you could fit a 100Mbit NIC or configure your switch to connect at 10/100Mbit instead of 1Gbit? Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Institute of Biological and Environmental Sciences, Cruickshank Building, St. Machar Drive, Aberdeen AB24 3UU, Scotland, UK. tel +44(0)1224 272700, fax +44 (0)1224 272 396 http://www.abdn.ac.uk, mailto:tony.travis at abdn.ac.uk, skype:ajtravis From tbooth at ceh.ac.uk Fri Jul 12 09:09:09 2013 From: tbooth at ceh.ac.uk (Tim Booth) Date: Fri, 12 Jul 2013 14:09:09 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: References: Message-ID: <1373634549.4263.30.camel@balisaur> Hi Alec, As Tony says, we need you to be a bit more specific about what you are doing on Bio-Linux that "hammers" the WAN. Downloading a large file on Bio-Linux should be no different to downloading a large file to any Windows machine, which I'm sure you do all the time, and I'd expect the network architecture to share bandwidth fairly without snarling up. It's common to rate-limit APT package updates because this is considered a low-priority operation, but if just downloading a file in Firefox, say, is snagging up the network then this would seem to be a more serious issue and possibly a network issue. In normal operation Linux is very well behaved in accessing the network - for example I run package updates at home over the broadband router and other devices can still access the internet. Cheers, TIM > > @TIM ? the problem is a single box hammering the WAN. I tried to > throttle the machine but it only effected the its LAN capabilities and > not the download (which is happening over the same connection). Is > there any way I can trace what is happening? > > > > Thanks for the help > > Alec > > > > -- Tim Booth NERC Environmental Bioinformatics Centre Centre for Ecology and Hydrology Maclean Bldg, Benson Lane Crowmarsh Gifford Wallingford, England OX10 8BB http://nebc.nerc.ac.uk +44 1491 69 2705 From tbooth at ceh.ac.uk Fri Jul 12 09:27:40 2013 From: tbooth at ceh.ac.uk (Tim Booth) Date: Fri, 12 Jul 2013 14:27:40 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: <51DFFF77.2090201@abdn.ac.uk> References: <51DFFC33.90704@abdn.ac.uk> <51DFFF77.2090201@abdn.ac.uk> Message-ID: <1373635660.4263.38.camel@balisaur> Hi, > Have you tried nice'ing the download program? > > nice -n 19 sftp ... "Nice" only affects CPU scheduling priority. It might help you to keep the ssh login responsive while the download is happening, but otherwise I can't see it doing much. Unless you only have a single CPU core in the server it won't make any difference at all. > To throttle at the hardware level, you could fit a 100Mbit NIC or > configure your switch to connect at 10/100Mbit instead of 1Gbit? This would seem a little desperate. I think you need to experiment with the -l/--limit-rate option to wget as Tony suggested. Maybe start at 1k/sec and try doubling the rate until you hit problems. Cheers, TIM -- Tim Booth NERC Environmental Bioinformatics Centre Centre for Ecology and Hydrology Maclean Bldg, Benson Lane Crowmarsh Gifford Wallingford, England OX10 8BB http://nebc.nerc.ac.uk +44 1491 69 2705 From tony.travis at abdn.ac.uk Fri Jul 12 09:34:32 2013 From: tony.travis at abdn.ac.uk (Tony Travis) Date: Fri, 12 Jul 2013 14:34:32 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: <1373635660.4263.38.camel@balisaur> References: <51DFFC33.90704@abdn.ac.uk> <51DFFF77.2090201@abdn.ac.uk> <1373635660.4263.38.camel@balisaur> Message-ID: <51E005E8.4090402@abdn.ac.uk> On 12/07/13 14:27, Tim Booth wrote: > Hi, > >> Have you tried nice'ing the download program? >> >> nice -n 19 sftp ... > > "Nice" only affects CPU scheduling priority. It might help you to keep > the ssh login responsive while the download is happening, but otherwise > I can't see it doing much. Unless you only have a single CPU core in > the server it won't make any difference at all. Hi, Tim. I know that :-) However, The way that "wget" and "sftp" limit bandwidth is by sleeping between blocks as the transfer proceeds. You're right that it won't make much difference on a system that is not limited by CPU resources. >> To throttle at the hardware level, you could fit a 100Mbit NIC or >> configure your switch to connect at 10/100Mbit instead of 1Gbit? > > This would seem a little desperate. Indeed... > I think you need to experiment with the -l/--limit-rate option to wget > as Tony suggested. Maybe start at 1k/sec and try doubling the rate > until you hit problems. I think this is good advice, but as Alec said before he already tried and if the way "wget" and "sftp" limit bandwidth is dependent on CPU resources being limited then it won't have much effect. These modern computers are just too fast for their own good ;-) Bye, Tony. -- Dr. A.J.Travis, University of Aberdeen, Institute of Biological and Environmental Sciences, Cruickshank Building, St. Machar Drive, Aberdeen AB24 3UU, Scotland, UK. tel +44(0)1224 272700, fax +44 (0)1224 272 396 http://www.abdn.ac.uk, mailto:tony.travis at abdn.ac.uk, skype:ajtravis From aleleb at sahfos.ac.uk Fri Jul 12 10:00:25 2013 From: aleleb at sahfos.ac.uk (Alec Colebrook-Clark) Date: Fri, 12 Jul 2013 15:00:25 +0100 Subject: [Bio-Linux] Bandwidth cont. In-Reply-To: <1373634549.4263.30.camel@balisaur> References: <1373634549.4263.30.camel@balisaur> Message-ID: Hi Tim and Tony, > As Tony says, we need you to be a bit more specific about what you are doing on Bio-Linux that "hammers" the WAN. All I am doing in this particular instance is downloading a collection of large files using 'wget' once I have ssh'd into the machine. I need to experiment more with the limit to see if it has the desired effect but initial tests aren?t looking great. I will have to try this more next week. It is worth noting that any package updates don?t cause an issue but I believe that is to do with their relatively small size. > It's common to rate-limit APT package updates because this is considered a low-priority operation, but if just downloading a file in Firefox, say, is >snagging up the network then this would seem to be a more serious issue and possibly a network issue. In normal operation Linux is very well >behaved in accessing the network - for example I run package updates at home over the broadband router and other devices can still access the >internet. I haven?t tried to see if the issue happens through a GUI. I normally just ssh in and run the download with 'wget'. We have other linux (virtual) machines and they are very well behaved. I shall discuss with the network manager to see if he can think of anything that might cause this. > However, The way that "wget" and "sftp" limit bandwidth is by sleeping between blocks as the transfer proceeds. You're right that it won't make >much difference on a system that is not limited by CPU resources. The machine isn?t limited on CPU but I shall look into it nonetheless. Thanks both for the advice, I have some further ideas now to try and tackle this problem! Alec -----Original Message----- From: bio-linux-bounces at nebclists.nerc.ac.uk [mailto:bio-linux-bounces at nebclists.nerc.ac.uk] On Behalf Of Tim Booth Sent: 12 July 2013 14:09 To: bio-linux at nebclists.nerc.ac.uk Subject: Re: [Bio-Linux] Bandwidth cont. Hi Alec, As Tony says, we need you to be a bit more specific about what you are doing on Bio-Linux that "hammers" the WAN. Downloading a large file on Bio-Linux should be no different to downloading a large file to any Windows machine, which I'm sure you do all the time, and I'd expect the network architecture to share bandwidth fairly without snarling up. It's common to rate-limit APT package updates because this is considered a low-priority operation, but if just downloading a file in Firefox, say, is snagging up the network then this would seem to be a more serious issue and possibly a network issue. In normal operation Linux is very well behaved in accessing the network - for example I run package updates at home over the broadband router and other devices can still access the internet. Cheers, TIM > > @TIM ? the problem is a single box hammering the WAN. I tried to > throttle the machine but it only effected the its LAN capabilities and > not the download (which is happening over the same connection). Is > there any way I can trace what is happening? > > > > Thanks for the help > > Alec > > > > -- Tim Booth NERC Environmental Bioinformatics Centre Centre for Ecology and Hydrology Maclean Bldg, Benson Lane Crowmarsh Gifford Wallingford, England OX10 8BB http://nebc.nerc.ac.uk +44 1491 69 2705 _______________________________________________ Bio-Linux mailing list Bio-Linux at nebclists.nerc.ac.uk http://nebclists.nerc.ac.uk/mailman/listinfo/bio-linux From tbooth at ceh.ac.uk Mon Jul 22 11:55:06 2013 From: tbooth at ceh.ac.uk (Tim Booth) Date: Mon, 22 Jul 2013 16:55:06 +0100 Subject: [Bio-Linux] Bio-Linux updated image and new packages Message-ID: <1374508506.10050.57.camel@balisaur> Hi All, I've put a new Bio-Linux image update on the website today (7.0.7). Existing users won't need to use this but I've added some new packages which you might be interested in installing on your machines (sudo apt-get install ): abyss (http://www.bcgsc.ca/platform/bioinfo/software/abyss) ray (http://denovoassembler.sourceforge.net/) sickle (https://github.com/vsbuffalo/sickle) scythe (https://github.com/vsbuffalo/scythe) macs (http://liulab.dfci.harvard.edu/MACS/) The R environment has been upgraded to R3.0 and modules have been recompiled to work on the new R. If you have installed extra modules and get an error about an incompatible R module you may need to run "install.packages()" within R. The new image has updated drivers from Ubuntu the 12.04.2 release and so should work on some more modern computers. I've not been able to test UEFI booting so if you have a UEFI machine you may need to set it back to BIOS boot for now. The new STADEN Gap5 package and Taverna did not make this image release but will be coming soon. Cheers, TIM -- Tim Booth NERC Environmental Bioinformatics Centre Centre for Ecology and Hydrology Maclean Bldg, Benson Lane Crowmarsh Gifford Wallingford, England OX10 8BB http://nebc.nerc.ac.uk +44 1491 69 2705 From tbooth at ceh.ac.uk Tue Jul 23 07:06:43 2013 From: tbooth at ceh.ac.uk (Tim Booth) Date: Tue, 23 Jul 2013 12:06:43 +0100 Subject: [Bio-Linux] Bio-Linux updated image and new packages In-Reply-To: <51ED59B7.9070600@gmail.com> References: <1374508506.10050.57.camel@balisaur> <51ED59B7.9070600@gmail.com> Message-ID: <1374577603.10050.91.camel@balisaur> Hi Andy, Yes, abyss is compiled with default K-mer size 64, as this can only be changed by recompilation of the binaries. A quick look at the code suggests that setting it to 96 would take a bit more memory for all jobs while setting it over 96 will cause the software to handle memory allocation in a completely different way (presumably resulting in a significantly larger memory footprint and/or slowdown). If you or other people on this list think that I should change the default in the package then let me know. With Velvet, I was able to make a package that contained extra versions of the executable compiled for different K-mer lengths but with the way ABYSS is set up (30 separate compiled executables and various scripts) this is not so easy. You can get the source quickly by doing: apt-get source abyss Then you can either recompile as per the usual instructions or look in debian/README.Debian for hints on how to recompile the DEB package with a different configuration. Thanks for the info on people who are using Bio-Linux. I've just come back from Berlin where a talk I gave seemed to go down pretty well and at least two people want to start using it for their own teaching, which is pretty cool. Cheers, TIM On Mon, 2013-07-22 at 17:11 +0100, Andrew Millard wrote: > Hi Tim, > > Just a quick question on these updates. Is the abyss package complied as > it would be if downloaded directly from the source so a default kmer > size of 64 ? If so am correct in thinking to change this, the source > would have be downloaded and recompiled to increase the kmer size as > there will be no config file in the package ? > > I am only asking, rather than testing myself as I am in the middle of > racking our new server that I will be installing biolinux on so dont > have anything to test it on > > > > Hope you are well > > cheers > andy > > > P.S I sent 2 visitors that I had spent a week learning bio-informatics > with me back to Germany with bio-linux USBs. They were most impressed > with the ease of use and are keen to use in there own research groups. > The Division I am now part of in the Medical School, all PhD students > and PDRAs are using bio-linux ( ~20 people) . I know having info like > this helps with justification of bio-linux etc , so thought I would > mention it > > From Andrew.Millard at warwick.ac.uk Tue Jul 23 07:30:33 2013 From: Andrew.Millard at warwick.ac.uk (Millard, Andrew) Date: Tue, 23 Jul 2013 11:30:33 +0000 Subject: [Bio-Linux] Bio-Linux updated image and new packages In-Reply-To: <1374577603.10050.91.camel@balisaur> References: <1374508506.10050.57.camel@balisaur> <51ED59B7.9070600@gmail.com>,<1374577603.10050.91.camel@balisaur> Message-ID: <77D2C3AE5CB9A944905CDB94C09E755078EE8764@AMSPRD0111MB500.eurprd01.prod.exchangelabs.com> Hi Tim, Thanks for the reply. I can see the issue with setting the Kmer setting upto 96. I hadn't realised that if compiled at 96 but only used a Kmer of 64 it meant an increase in memory usage compared to compiling to at 64 Kmer. I have downloaded and compiled from source at a higher Kmer and it is a relatively painless task and matter of following their online instructions. I personally think that the having a default maximum Kmer of 96 would be better as with the increase in reads of upto 250 bp then having the ability to use a kmer of upto 96 "out of the box" will give "better" assembly. My experience from doing bacterial assemblies is that programs such as Kmergenie often predict the optimal Kmer is above 64, so would be advantageous to have the ability to use higher Kmer. andy Dr Andrew Millard Division of Microbiology and Infection Warwick Medical School The University of Warwick Coventry CV4 7AL http://www2.warwick.ac.uk/fac/sci/lifesci/people/amillard Tel 0247615045 ________________________________________ From tbooth at ceh.ac.uk Tue Jul 23 10:08:12 2013 From: tbooth at ceh.ac.uk (Tim Booth) Date: Tue, 23 Jul 2013 15:08:12 +0100 Subject: [Bio-Linux] Bio-Linux updated image and new packages In-Reply-To: <77D2C3AE5CB9A944905CDB94C09E755078EE8764@AMSPRD0111MB500.eurprd01.prod.exchangelabs.com> References: <1374508506.10050.57.camel@balisaur> <51ED59B7.9070600@gmail.com>,<1374577603.10050.91.camel@balisaur> <77D2C3AE5CB9A944905CDB94C09E755078EE8764@AMSPRD0111MB500.eurprd01.prod.exchangelabs.com> Message-ID: <1374588492.10050.135.camel@balisaur> Hi Andy, > Thanks for the reply. > I can see the issue with setting the Kmer setting upto 96. I hadn't > realised that if compiled at 96 but only used a Kmer of 64 it meant an > increase in memory usage compared to compiling to at 64 Kmer. It's hard to see how big a difference this will make in practise. I'd need to run some test data on both versions and see how it actually impacts the memory usage. > I have downloaded and compiled from source at a higher Kmer and it is > a relatively painless task and matter of following their online > instructions. Yes, the thing compiles quite happily, so the Bio-Linux package doesn't add much aside from a bit of convenience. > I personally think that the having a default maximum Kmer of 96 would > be better as with the increase in reads of upto 250 bp then having the > ability to use a kmer of upto 96 "out of the box" will give "better" > assembly. My experience from doing bacterial assemblies is that > programs such as Kmergenie often predict the optimal Kmer is above 64, > so would be advantageous to have the ability to use higher Kmer. OK, I'm convinced. Some people are always going to need to need to recompile, as some people will want a lower-memory version or super-long Kmers, but it's more important that the default version can deal with typical data. I'll modify the default in the package. Cheers, TIM > Dr Andrew Millard > Division of Microbiology and Infection > Warwick Medical School > The University of Warwick > Coventry > CV4 7AL > http://www2.warwick.ac.uk/fac/sci/lifesci/people/amillard > Tel 0247615045 > ________________________________________ > > _______________________________________________ > Bio-Linux mailing list > Bio-Linux at nebclists.nerc.ac.uk > http://nebclists.nerc.ac.uk/mailman/listinfo/bio-linux -- Tim Booth NERC Environmental Bioinformatics Centre Centre for Ecology and Hydrology Maclean Bldg, Benson Lane Crowmarsh Gifford Wallingford, England OX10 8BB http://nebc.nerc.ac.uk +44 1491 69 2705