ILRI Research Computing

User Tools

Site Tools

You are here: start » hpc
hpc

This is an old revision of the document!

Changelog

qrsh_command                 /usr/bin/ssh
rsh_command                  /usr/bin/ssh
rlogin_command               /usr/bin/ssh
  • Created link from /tmp → /state/partition1 on the head node as this is the "temp" space to be used on the nodes, and it needs to be available on the head node as well.
$ sudo mkdir -p /state/          
$ sudo ln -s /scratch/ /state/partition1
$ sudo chmod 777 /scratch/
$ sudo chmod o+t /scratch/
  • January 20, 2010: Installed 32-bit compatibility libraries yum install compat-libstdc++
  • January 22, 2010:
    • Removed 32-bit packages from HPC: 32bit
    • Ran some trivial application updates: # yum update wget vim-common vim-enhanced vim-minimal xterm tcsh sudo rpm rpm-libs rpm-devel openmpi\* gzip grep less ksh unix2dos rsh readline mdadm elinks gcc\* evince curl strace screen mysql mysql-server mlocate nash setup graphviz graphviz-perl
  • January 26, 2010: Ran trivial application updates: # yum update vino tzdata tftp-server tcl tcl-devel python python-devel mpi-selector metacity lftp gd firefox gd-devel ethtool cpio coreutils gnutls ghostscript cvs centos-release
  • February 23, 2010: Disabled GSSAPIAuthentication in /etc/ssh/sshd_config because we don't use Kerberos and it was causing lag on authentication.
  • March 4, 2010: Added @users - priority 19 to /etc/security/limits.conf to set the process priority of users in the "users" group to 19 (-20 is the highest priority, +19 is the lowest).
  • March 5, 2010: Added options to /etc/ssh/ssh_config to enable X forwarding (-X AND -Y, insecure and secure) for SSH connections from HPC:
ForwardX11              yes
ForwardX11Trusted       yes
  • March 9, 2010: Changed the SGE configuration (qconf -mconf) to allow for X forwarding while logging into a random node using qrsh:
rsh_command                  /usr/bin/ssh
rlogin_command               /usr/bin/ssh
rsh_daemon                   /usr/sbin/sshd -i
rlogin_daemon                /usr/sbin/sshd -i
qlogin_command               builtin
qlogin_daemon                builtin

See http://gridengine.sunsource.net/howto/qrsh_qlogin_ssh.html

  • March 19, 2010: Installed B2G4PIPE (blast2go- A version for B2G annotation without FrontEnd (GUI) for Pipeline Integration) http://www.blast2go.de/ ; Created a script 'blast2go' script to load 'java -jar blast2go.jar' can be accessed on all the nodes
  • March 30, 2010: Installed Staden packages 1.7 and 2.0.0b6 to /paracel/paracel/biosoft/staden/, and added the 2.0.0b6 version to the PATH. 1.7 needed some links to old libraries in order to function:
$ cd /paracel/paracel/biosoft/staden/staden-linux-x86_64-1-7-1b/lib/linux-x86_64-binaries/
$ sudo ln -s /lib64/libssl.so.0.9.8e libssl.so.0.9.7
$ sudo ln -s /lib64/libcrypto.so.0.9.8e libcrypto.so.0.9.7
  • April 8, 2010: Archived trushar's data to a tape because it was 110 gigs:
$ tar -cpzvf /dev/st0 /mnt/export/home/trushar

Note, it is gzipped so you must use tar zxvf to unzip it: 

$ tar zxvf /dev/st0 mnt/export/home/trushar/
  • April 12, 2010: Installed gsRunBrowser 2.3 from the FLX cluster, somehow this one slipped through when we copied the new applications from the FLX in December.
  • April 20, 2010: Installed PhyML 3.0 to /usr/local/bin/PhyML_3.0_linux64. It should soon replace PHYLIP.
  • April 21, 2010: Installed latest PHYLIP to /paracel/paracel/biosoft/phylip-3.69, and modified the PATH in /etc/profile.d/paracel.sh to include it instead of the older 3.67 version. It should be noted, however, that Rocks also includes a PHYLIP package, version 3.68.
  • April 23, 2010: Installed a 32-bit version of MrBayes due to an apparently known bug with 64 bit versions and some data sets crashing during sumt/sump operations. It's available here: /paracel/paracel/biosoft/mrbayes-3.1.2/mb32
  • May 12, 2010: Ran trivial software updates:
# yum install tar coreutils.x86_64 cpio firefox.x86_64 libXi.i386 libXi.x86_64 libvorbis.x86_64 libxml2.x86_64 \
libxml2-python.x86_64 ntp.x86_64 openssh.x86_64 sudo.x86_64 strace perl tcp_wrappers.x86_64 tzdata.noarch \
openssl.x86_64 openssh-clients.x86_64 python26.x86_64 openssl-devel.x86_64 openssh-server.x86_64  sqlite.x86_64 \
sqlite-dev.x86_64 psmisc.x86_64 pciutils.x86_64 pciutils-dev.x86_64 iptables.x86_64 dhclient.x86_64 \
chkrootkit.x86_64

It's easy to run trivial software updates after looking at the output to the yum list updates command and picking non-critical software (things which don't need a reboot).

  • May 13, 2010: Copied some missing files to the compute nodes… parse the output of lsload to find hosts which are up, then use a bash loop to rsync the files:
for host in $(lsload | grep ok | grep -v hpc-ilri | awk '{print $1}' | sed 's/\..*$/.local/');  do echo "About to copy files to ${host}..."; rsync -av --delete /opt/Bio/fasta/ ${host}:/opt/Bio/fasta/; done
  • May 18, 2010: Install Biopython and NumPy for python2.6
  • May 18, 2010: Head node crashed after a heavy python script was run (using python 2.6 and Biopython) so we used it as an opportunity to upgrade several packages which require a reboot since there was downtime anyways (kernel, udev, hal, etc).
  • May 18, 2010: Installed a new Python 2.6.5 to /paracel/paracel/biosoft/python/2.6.5 for use by the nodes. Also installed Biopython.
  • May 20, 2010: Downloaded EagleView to /paracel/paracel/biosoft/eagleview.
  • May 22, 2010: Install the environment modules roll. See modules
  • May 25, 2010: Several ports have been allowed to the HPC: SSH, NTP, ICMP.
  • Limit SSH access to users in the ssh group.
  • Limit compiler access to users in the gcc group.
  • June 4, 2010: Enable PermitRootLogin without-password in /etc/ssh/sshd_config so the backup scripts from other servers can run as root with key-based authentication.
  • June 22, 2010: Installed some software for the Phylogenetics course:
    • BEAST
    • RAxML
  • June 29, 2010: Installed more software for the Phylogenetics course:
    • Dendroscope 2.7.3
    • Dendroscope 2.7.4
    • Splitstree 4.11.3
  • June 30, 2010: Installed the Celera Whole Genome Shotgun Assembler and appropriate module files:
    • /paracel/paracel/biosoft/celera-wgs/6.1
  • August 2, 2010: Change the SGE queue to all.q for Workflow (in /mnt/export/paracel/biosoft/workflow/3.1.2/bin/sge_submitter.sh). The default was workflow.q.
  • Add aorth as a manager to SGE so he can use qmon without root (there are problems running as root anyways):
    • qconf -am aorth
  • August 10, 2010: Updates from yum:
    • # yum update lftp.x86_64 ksh.x86_64 nash.x86_64 p7zip.x86_64 p7zip-plugins.x86_64 parted.x86_64 rpmforge-release.x86_64 rsync.x86_64 rtorrent.x86_64 subversion.x86_64 chkrootkit.x86_64 automake.noarch dosfstools.x86_64 evince.x86_64  ethtool.x86_64 gnutls.x86_64 kpartx.x86_64
  • August 12, 2010: Edit R script to use SGE for picking a random node, /paracel/paracel/biosoft/R/2.11.0/bin/R:
    • #qrsh xterm -e $PROGRAM_CMD
qrsh $PROGRAM_CMD --save
    • The commented out way requires that R runs via an xterm (and fails if X11 isn't active).
  • September 8, 2010:
    • Fine tune SSH access to HPC. The ssh daemon only accepts connections from people in the ssh group, but we don't want some people to log in from external to ILRI. Added this to /etc/security/access.conf:
      • # default rules for root
+   :   root        :   crond :0 tty1 tty2 tty3 tty4 tty5 tty6
+   :   root        :   127.0.0.1

# some special accounts who can access from anywhere
+   :   aorth       :   ALL 
+   :   spatel      :   ALL 
+   :   evilliers   :   ALL 
+   :   afischer    :   ALL 

# everyone else from the corporate network: 192.168.1.2
+   :   ALL         :   192.168.1.2/255.255.255.255

# deny everything else
-   :   ALL         :   ALL
    • This allows root locally, a few people from ALL, then everyone else from only 192.168.1.2.
  • September 19, 2010:
    • It appears HPC crashed this weekend… the hardware RAID hiccuped or something, so I had to reboot the server. I had thought maybe a drive died, but after reboot everything was OK.
    • I took that opportunity to run some system updates as well: kernel, openssh-server, libpng, udev, glibc, sudo, and a few others which looked like they wouldn't cause problems but were probably important…
  • September 30, 2010:
    • Updated the zabbix agent:
      • # yum update zabbix.x86_64 zabbix-agent.x86_64
    • Still not working for monitoring, need to add a zabbix user and configure the agent…
  • October 1, 2010:
    • Couldn't get the zabbix agent to work, seems there have been several installs/configurations.
    • Deleted all zabbix-related stuff and reinstalled to start from scratch:
      • # rm -rf /etc/zabbix
# rm -rf /var/log/zabbix*
# rm -rf /var/run/zabbix*
# rm -rf /usr/share/zabbix
# yum remove zabbix zabbix-agent
# yum install zabbix-agent
    • Update /etc/zabbix/zabbix-agentd.conf:
      • Server=192.168.1.2
PidFile=/var/run/zabbix/zabbix_agentd.pid
DisableActive=1
    • Create the zabbix run directory to store the pid:
      • # mkdir /var/run/zabbix/
# chown -R zabbix /var/run/zabbix
    • Start the zabbix agent:
      • # /etc/init.d/zabbix-agentd start
  • October 12, 2010:
    • Move SSH-specific PAM access rules to the appropriate file, /etc/security/sshd_access.conf.
    • I had mistakenly added some to /etc/security/access.conf and wondered why they didn't work.
    • See the configuration of pam_access related to SSH in /etc/pam.d/sshd
    • Also, comment out the root rules in /etc/security/access.conf because they are not necessary (oops).
  • October 19, 2010:
    • Power problems at ILRI, had to rebuild the nodes. From the head node, set their boot and install actions to appropriate settings:
    • rocks set host boot compute action="install"
rocks set host installaction compute action="install ilri"
    • Then physically reboot them. They will PXE boot to their installers…
  • November 8, 2010:
    • Install bowtie, samtools, python 2.6.6, and corresponding module files.
    • Add /mnt/export/paracel/biosoft/bin to the system's PATH in /etc/profile.d/paracel.sh. I want to make it easier to use certain applications without appending their paths to the PATH and also not force the use of environment modules for every single app we install. Hopefully this will be the beginning of cleaning up the hideous PATH.
      • Find directories in the PATH which don't exist anymore, crude one-liner:
      • # for dir in $(head -n 1 /etc/profile.d/paracel.sh | sed 's/:/\n/g'); do test -d "$dir"; if [ $? -eq 1 ]; then echo $dir; fi; done
PATH=$PATH
/paracel/paracel/biosoft/codonw
/usr/local/PolyBayes/bin
/opt/mpich/bin
      • … and remove them from the PATH.
      • Move the existing /mnt/export/paracel/biosoft/bin to oldbin as it looks like it hasn't been used in 5 years. Can delete later when we're sure nothing depends on it…?
      • Link a few binaries into the new bin folder for easy access:
      • # ln -s /mnt/export/paracel/biosoft/bwa/0.5.8c/bwa .
# ln -s /mnt/export/paracel/biosoft/samtools/0.1.9/samtools .
# ln -s /mnt/export/paracel/biosoft/bowtie/0.12.7/bowtie .
# ln -s /mnt/export/paracel/biosoft/bowtie/0.12.7/bowtie-build .
# ln -s /mnt/export/paracel/biosoft/bowtie/0.12.7/bowtie-inspect .
      • Now these binaries can be easily accessed without having to use environment modules. Environment modules are really more suited for packages with lots of data, man pages, libs, etc (like Python, etc). We can replace the links with links to newer versions of binaries as they are released.
      • One bonus of this method seems to be that the nodes can access them… seems to work well?
  • November 18, 2010:
    • Add excludes to headnode's yum.conf:
    • exclude=dapl* ibutils* infiniband* kernel-ib* libibumad* libibmad* libibverbs* libmthca* libibverbs* libnes* librdmacm* ofed* qperf* libmlx4* compat-dapl* opensm* libcxgb3* libibcom* openib* openmpi* ibvexdmtools* libibpathverbs* libipathverbs* libsdp* srptools* ibsim*
  • November 19, 2010:
    • Install mira 3.2.0rc2, and remove older versions.
  • November 21, 2010:
    • Power failure on HPC this weekend, maybe someone shut it down because of the problems in the server room? Still not sure…
    • Ran the following updates:
    • $ yum update sudo.x86_64 perl\* python\* pam\* kernel\*
    • Noticed a problem with the /etc/hosts.allow file, not sure who added it so I'm commenting it out:
    • #172.26.15.184
  • November 23, 2010:
    • Replace CMOS in 8 nodes, should help with nodes not remembering boot order(PXE first!) after power failure.
  • November 26, 2010:
    • Install consed 20.0
    • $ sudo mkdir -p /mnt/export/paracel/biosoft/consed/20.0/
$ sudo mkdir -p /mnt/export/paracel/biosoft/consed/20.0/bin/
$ sudo cp consed_linux64bit /mnt/export/paracel/biosoft/consed/20.0/bin/
$ sudo ln -s /mnt/export/paracel/biosoft/consed/20.0/bin/consed_linux64bit /mnt/export/paracel/biosoft/consed/20.0/bin/consed
$ cd misc/phd2fasta
$ make
$ sudo cp phd2fasta /paracel/paracel/biosoft/consed/20.0/bin/
$ cd ../mktrace
$ make
$ sudo cp mktrace /paracel/paracel/biosoft/consed/20.0/bin/
$ cd ../454/
$ gcc sff2scf.c -o sff2scf
$ gcc sff2scfAndPhd.c -o sff2scfAndPhd
$ gcc sffinfo.c -o sffinfo
$ sudo cp sff2scf sff2scfAndPhd sffinfo /paracel/paracel/biosoft/consed/20.0/bin/
$ cd ../../scripts
$ sudo cp * /paracel/paracel/biosoft/consed/20.0/bin/
$ sudo mkdir /paracel/paracel/biosoft/consed/20.0/lib
$ sudo cp /mnt/export/paracel/biosoft/phred/phredpar.dat /paracel/paracel/biosoft/consed/20.0/lib/
$ sudo mkdir /paracel/paracel/biosoft/consed/20.0/lib/screenLibs/
$ cd ../misc/
$ sudo cp *.fa *.fasta *.seq /paracel/paracel/biosoft/consed/20.0/lib/screenLibs/
      • I disabled 454's sff2scf as per consed's README.txt
      • Created a module file for consed
    • January 17, 2010:
      • # yum update xorg-\* vnc-\* util-linux tetex-\* tcsh python-\* postgresql-\* xulrunner php\* p7zip\* openssl\* openldap\* nfs-utils\* nash mysql\* lib\* kernel\* glib\* firefox\* gtk\* iptables\* freetype evince e2fsprogs\* dhclient expat\* dhcp\* chkconfig ImageMagick apr\*
      • HPC was shutdown in the weekend due to a power outage, so it was a good chance to reboot it anyways before too many people started running jobs.
        • Still haven't had time to turn on all the nodes.
  • January 20, 2011:
    • Updated 454 applications to 2.5.3 (gsAssembler, etc):
      • the official installers put the files in /opt/454 but after installation I moved them back to /paracel/paracel/biosoft/454
  • January 28, 2011:
    • Install ARB in /paracel/paracel/biosoft/arb/5.2 for Anne Fischer: http://www.arb-home.de
    • HPC crash because of Anne's ARB… had to sysrq…
  • February 6, 2011:
    • Had to restart lsf on the head and compute nodes to fix the "missing license" problem.
  • February 11, 2011:
    • Resurrect some compute nodes by fixing their grub parameters at boot: nomce tsc needs to be there or else they don't boot!
    • Shut some compute nodes down to save heat during the weekend's air conditioning service
  • February 14, 2011:
    • Turn compute nodes back on
    • Install mira 3.2.1, /paracel/paracel/biosoft/mira/3.2.1
  • February 18, 2011:
    • Install marthlab's Unified Release software package (gigaBayes, Mosaik, etc) in /paracel/paracel/biosoft/marthlab
    • Install Strand's AvadisNGS (next-generation sequencing tools) but didn't license it yet because the demo only lasts 20 days. /paracel/paracel/biosoft/strand/AvadisNGS
  • March 31, 2011:
    • Reboot HPC because some problems with the KVM's mouse, and the tape drive is behaving funny
    • Update Storix SBadmin from 6.3.4.4 to 6.3.4.5
    • Trying to clean up the log files, drop packets from CIP's HPC trying to grid with us:
    • sudo iptables -I INPUT -s 200.62.229.37 -j DROP
  • May 3, 2011:
    • Run updates:
    • sudo yum update yum.noarch tzdata vim-\* vnc-\* sudo subversion screen tmux python\* perl-\* rpm pcre\* patch passwd p7zip\* rpmforge-release.x86_64 mysql\* mtx nfs-utils metacity mesa\* openssh\* lib\* logrotate man-pages psmisc logwatch httpd\* autofs
  • May 5, 2011:
    • Reconfigure the NFS server's ports to be static, /etc/sysconfig/nfs:
    • LOCKD_TCPPORT=32803
LOCKD_UDPPORT=32769
MOUNTD_PORT=892
RQUOTAD_PORT=875
STATD_PORT=662
STATD_OUTGOING_PORT=2020
    • Reconfigure iptables to allow connections from the corporate ILRI network, /etc/sysconfig/iptables:
    • # NFS for ILRI corporate network
-A INPUT -s 192.168.1.2/32 -m state --state NEW -p udp --dport 111 -j ACCEPT
-A INPUT -s 192.168.1.2/32 -m state --state NEW -p tcp --dport 111 -j ACCEPT
-A INPUT -s 192.168.1.2/32 -m state --state NEW -p tcp --dport 2049 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p tcp --dport 32803 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p udp --dport 32769 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p tcp --dport 892 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p udp --dport 892 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p tcp --dport 875 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p udp --dport 875 -j ACCEPT
-A INPUT -s 192.168.1.2/32  -m state --state NEW -p tcp --dport 662 -j ACCEPT
-A INPUT -s 192.168.1.2/32 -m state --state NEW -p udp --dport 662 -j ACCEPT
  • May 11, 2011:
    • Reboot HPC again and REALLY apply the updates from May 3. Appears they didn't get run last time.
  • May 16, 2011:
    • Update SSH access configuration in /etc/security/sshd_access.conf to reflect new ICT network configuration:
    • # everyone else from the corporate network: 192.168.5.1
+   :   ALL         :   192.168.5.1/255.255.255.255
  • May 17, 2011:
    • Re-enable the system MySQL. We had disabled it and moved databases to the Rocks MySQL but this is kinda annoying when you start having perl modules, php, etc which are configured to use the system's MySQL socket.
  • May 27, 2011:
    • Hmmm, for some reason the Rocks MySQL wasn't running on HPC. Start it:
    • sudo service foundation-mysql start
  • May 31, 2011:
    • Install new version of clustalw into /paracel/paracel/biosoft/clustalw/2.1
    • Update module files for R, python, clustalw, and dendroscope to use a cleaner, more-robust syntax for conflicts and man pages

New HPC!

  • June 18, 2011:
    • Install new HPC!
  • June 20, 2011:
    • Install muscle 3.8.31 into /usr/local/bin (http://www.drive5.com/muscle/)
    • IPtables rules for samba seem to need some work, windows clients are having problems connecting so I've disabled iptables for now.
  • June 28, 2011:
    • Install phred, phrap, and cap3 + module files.
  • June 30, 2011:
    • Install seqclean, estscan, and tgcil + module files. They are part of the est2uni package, which we haven't installed yet! est2uni should have a module file which loads them as dependencies.
  • July 1, 2011:
    • Remove the DefaultTime from the slurm batch partition… it's too hard for users to know how long their jobs will run.
    • Allow aorth to get the status of the hardware RAID without a password, /etc/sudoers:
    • Cmnd_Alias RAIDSTATUS = /opt/MegaRAID/MegaCli/MegaCli64 -PDList -aALL
# allow Alan to query the RAID status remotely.
aorth   ALL = NOPASSWD: RAIDSTATUS
    • Remove Rocks' Wordpress because Ian Moore said it showed up in an external security scan and we're not using it.
      • sudo yum remove wordpress
  • July 5, 2011:
    • Make changes to allow the RAID monitoring script to run via a cron job (disallow the "requiretty" option) /etc/sudoers:
    • # Allow the RAIDSTATUS command alias to use sudo from a cron job remotely (without a tty!... for the RAID monitoring script)
# see `man sudoers` cuz this syntax is confusing.
Defaults!RAIDSTATUS !requiretty
  • July 12, 2011:
    • Change php's mysqli and mysql sockets to the system mysql (they were set to Rocks'!) and it was messing up some stuff
    • Installed formtools
    • Remove the pre-installed mediawiki as it's just going to get out of date!
    • Install VirtualBox 4.0.10:
    • wget http://download.virtualbox.org/virtualbox/debian/oracle_vbox.asc
sudo rpm --import oracle_vbox.asc
wget http://download.virtualbox.org/virtualbox/rpm/rhel/virtualbox.repo
sudo cp virtualbox.repo /etc/yum.repos.d/
sudo yum list available virtual\*
sudo yum install VirtualBox-4.0.x86_64
  • July 21, 2011:
    • Interproscan done installing. CLI tests work, but lets see if the web frontend works: http://hpc.ilri.cgiar.org/iprscan/iprscan
    • Installed glimmer3 + module file but it seems Rocks already has a glimmer3… hmmm.
  • August 4, 2011:
    • Installed Affymetrix Powertools (APT) in /export/apps/apt/1.14.3
    • Installed updates: yum update VirtualBox-4.0.x86_64 curl curl-devel gnome-terminal.x86_64 ksh.x86_64 openmotif openmotif-devel ypbind.x86_64
    • Update VirtualBox to 4.1.0
  • August 5, 2011:
    • Fix the lims server's automatic backups to HPC…
      • Add root user to the ssh group
      • Edit /etc/security/sshd_access.conf:
      •  # root from the corporate network: 192.168.5.1
+   :   root            :192.168.5.1/255.255.255.255
  • August 10, 2011:
  • August 11, 2011:
    • Enable java plugin on Firefox:
    • sudo ln -s /usr/java/jdk1.6.0_16/jre/lib/amd64/libnpjp2.so /usr/lib64/mozilla/plugins/
  • August 22, 2011:
    • Install boost 1.47.0, samtools 0.1.17, and cufflinks 1.0.3
    • Write documentation and module files for each
    • Update galaxy to use newer samtools (0.1.17 instead of 0.1.16)
  • September 1, 2011:
    • Install MEGAN:
      • /export/apps/megan/4.60.2
      • set the default memory allocation to 16G in /export/apps/megan/4.60.2/MEGAN.vmoptions (default was 2000M)
  • September 2, 2011:
    • Install kernel updates
    • Install new SAS card to prepare for tape library
    • Enable virtualization support in the BIOS
    • Apparently MySQL and munge weren't set to start during boot (causing various dependent services to fail), so enable them:
    • sudo chkconfig --level 2345 mysqld on
sudo chkconfig --level 2345 munge on
    • Install apcupsd but I'm actually not sure if the power is coming from the UPS right now…
  • September 6, 2011:
    • Fix the module files for est2uni and estscan (they were using setenv to set the PERL5LIB variable, and overwriting eachother! They needed to use prepend-path instead)
    • Update CrossmatchSearchEngine.pm to fix a minor bug in RepeatMasker (more info here: http://repeatmasker.systemsbiology.net/pipermail/rmannounce/2011-June/000060.html)
    • Install rmblast 2.2.23+ in /export/apps/rmblast/2.2.23+ for use by RepeatMasker and changed RepeatMasker to use it by default (as cross_match seems to be giving some errors)
  • September 22, 2011:
    • Install the new tape library (but haven't set up software yet!)
  • September 23, 2011:
    • Install mira 3.4.0 (and module file) in /export/apps/mira/3.4.0
    • Install staden 2.0.0b8 (and module file) in /export/apps/staden/2.0.0b8
  • September 27, 2011:
    • Install clview:
    • mkdir clview
tar zxf clview_linux_i386.tar.gz -C clview
cd clview
sudo cp libFOX-1.0.so.0* /usr/lib/
sudo cp clview /usr/local/bin/
sudo chmod 755 /usr/lib/libFOX-1.0.so.0* /usr/local/bin/clview
clview
  • September 29, 2011:
    • Add Harry Noyes to the ssh group on HPC, and to /etc/security/sshd_access.conf so he can transfer some files for Sheila.
    • Access to GFHPC and HPC from outside is still down though, James Indimuli in ICT says he'll get back to me.
hpc.1317305841.txt.gz · Last modified: by aorth