Skip to content

Instantly share code, notes, and snippets.

View LincolnBryant's full-sized avatar
💭
Input translation, output rotation

Lincoln Bryant LincolnBryant

💭
Input translation, output rotation
View GitHub Profile
Installing dcache 2.6.x from scratch
====================================
Installing Java 7
-----------------
Grab the Java 7 RPM:
````
[root@dfs-m1 ~]# yum localinstall http://bootstrap.mwt2.org/repo/MWT2/jre-7-linux-x64.rpm
````

Installing from scratch

Starting from a fresh install of EL6.5, first install EPEL:

# rpm -Uvh http://download.fedoraproject.org/pub/epel/6/i386/epel-release-6-8.noarch.rpm

Then install the Graphite dependencies:

# yum -y install python-whisper python-carbon graphite-web python-memcached python-ldap httpd memcached nodejs npm nc

Tahoe-LAFS Build for CentOS 6.5

requirements

yum install openssl-devel libffi-devel

creating virtualenv

virtualenv tahoe
git clone https://github.com/tahoe-lafs/tahoe-lafs
@LincolnBryant
LincolnBryant / gist:7161bf402c8103acd0b6
Created September 10, 2014 05:56
fast & loose build reqs for Tahoe on CentOS / Scientific Linux 6.5

Tahoe-LAFS Build for CentOS 6.5

requirements

yum install openssl-devel libffi-devel

creating virtualenv

@LincolnBryant
LincolnBryant / gist:c0d1a9e198b77ca8a76a
Last active August 29, 2015 14:06
Tahoe-LAFS build & setup under CentOS 6.5

Tahoe-LAFS Build for CentOS 6.5

build requirements

yum install openssl-devel libffi-devel

creating virtualenv

unset _CONDOR_JOB_PIDS
unset _CONDOR_ANCESTOR_6635
unset _CONDOR_ANCESTOR_1130
unset _CONDOR_SCRATCH_DIR
unset _CONDOR_ANCESTOR_1092
unset _CONDOR_CHIRP_CONFIG
unset _CONDOR_ANCESTOR_1095
unset _CONDOR_WRAPPER_ERROR_FILE
unset _CONDOR_SLOT
unset _CONDOR_EXECUTE
StashCache=(`grep ^WantsStashCache $_CONDOR_JOB_AD`)
PosixStashCache=(`grep ^WantsPosixStashCache $_CONDOR_JOB_AD`)
function setup_stashcp {
module load xrootd
module load stashcp
# Determine XRootD plugin directory.
# in lieu of a MODULE_<name>_BASE from lmod, this will do:
export MODULE_XROOTD_BASE=$(which xrdcp | sed -e 's,/bin/.*,,')

the libvirt netinstall bargain bonanza!

I think in my ideal world, I'd have unlimited hardware to install every quirky Linux or UNIX distro under the Sun. Since I don't live in that glorious fantasy realm, I've taken to virtualizing pretty much everything lately. One of the things that irritates me about that approach, though, is that I have to download a relatively fat ISO in order to get at a fairly small installer and bootstrap my shiny new OS.

The "Campus Factory" script basically monitors schedds for idle jobs and submits as appropriate to the remote cluster. A typical submission looks like this:

condor_submit /home/ruc.clemson/bosco/libexec/campus_factory/share/glidein_jobs/job.submit.template  -a REMOTE_SCHEDULER="pbs" -a WN_TMP="/local_scratch" -a GLIDEIN_Site="osgconnect@user.palmetto.clemson.edu/pbs" -a GLIDEIN_DIR="/home/ruc.clemson/bosco/libexec/campus_factory/share/glidein_jobs" -a REMOTE_FACTORY="/scratch1/osgconnect/bosco/clemson/campus_factory" -a PASSWDFILE_LOCATION="/home/ruc.clemson/bosco/local.bosco/passwdfile" -a BOSCOCluster="osgconnect@user.palmetto.clemson.edu/pbs" -a GLIDEIN_HOST="rccf-osg.ci-connect.net:11012?sock=collector" -a REMOTE_CLUSTER="osgconnect@user.palmetto.clemson.edu"

I suspect APF can completely replace this component.

The job submit template looks like this:

#!/bin/bash
################################################################################
# blahp-over-ssh shim script
#
# command [options] remote_hostname [remote options/arguments]
#
# Options:
# see usage string
#