SIO Paleomag Lab blog

The sample tower for BUBBA consists of the main base, a custom fabricated sample holder, and a lowering system set up with fishing line. The fishing line is always streching and eventually fails. Well, today it snapped while Julie was using it and she restrung the tower with fresh fishing line. There is a new diagram on how to restring the tower just to the north of BUBBA. It usually takes a couple of attempts to have the fishing line tight enough so it doesn't slip along the wheel. Hopefully, this set up will last forever. :)

Today, Julie was going to show me how to use the curie balance. When we started going through the startup procedure, we discovered that water had been leaking from the cooling loop. We shutoff the cooling lines right behind the chiller, but if you use the AGFM, make sure that you check for water leaking from the lines right where they go into the furnace and the two magnets. Jason mentioned that the furnace had a bad weld the last time there was a leak, but this time it looks like a leak at both the furnace and the front magnet.

This script logs onto each computer, makes a tar of the RAWDAT files and sends a bzip2 back to Sorcerer.

#!/bin/bash
## This program logs onto each computer and
makes a tar of the RAWDATA
## The tar file is then zipped and copied back to Sorcerer
## Must be logged in as pmag or else!

LAB=/Desktop/LabVIEW-Data-`date "+%C%y%m%d"`.tar
MINI=/Desktop/MINISPIN-RAWDAT-`date "+%C%y%m%d"`.tar
AMS=/Desktop/KAPPABRIDGE-RAWDAT-`date "+%C%y%m%d"`.tar

echo "MAKING TAR ON PIXIE"
ssh pixie "tar cf ~$LAB ~/Documents/LabVIEW\ Data/"
echo "ZIPPING TAR ON PIXIE"
ssh pixie "bzip2 -9 ~$LAB"
echo "TRANSFERING TO SORCERER"
scp pixie:~$LAB.bz2 ~/Desktop/TAR-BZ2
ssh pixie "cd ~/Desktop ; rm ~$LAB.bz2"

echo "MAKING TAR ON ANORTHITE"
ssh anorthite "tar cf ~$MINI ~/Documents/MINISPIN-RAWDAT"
echo "ZIPPING TAR ON ANORTITE"
ssh anorthite "bzip2 -9 ~$MINI"
echo "TRANSFERING TO SORCERE"R
scp anorthite:~$MINI.bz2 ~/Desktop/TAR-BZ2
ssh anorthite "cd ~/Desktop ; rm ~$MINI.bz2"

echo "MAKING TAR ON SOOTHSAYER"
ssh soothsayer "tar cf ~$AMS ~/Documents/KAPPABRIDGE-RAWDAT"
echo "ZIPPING TAR ON SOOTHSAYER"
ssh soothsayer "bzip2 -9 ~$AMS"
echo "TRANSFERING TO SORCERER"
scp soothsayer:~$AMS.bz2 ~/Desktop/TAR-BZ2
ssh soothsayer "cd ~/Desktop ; rm ~$AMS.bz2"

The trick to this was being able to log onto the computers without being asked for a password so backups can be run through a crontab. The command to do this is

ssh-kegen -t dsa

This creates public/private keys in the .ssh/. The id_dsa.pub is your public key. This file has to be copied into .ssh/ of the machine which it wants to log into and called authorized_keys. So, when Sorcerer ssh's into select computer's, it's fingerprint is already known through an authorized_key and you are not asked for a password.

Here are some pictures from July 2004 that show the insides of Bubba the magnetometer along with many of the people in the SIO Paleomagetics Lab.

  Bubba's Guts and Fill

A liquid helium fill was performed on BUBBA today and it wasn't the smoothest fill. To begin with, the liquid helium supply dewar had 38 liters recorded on the tag, though we measured only 28 liters. Then, when we began to fill BUBBA the exhaust plume appeared weaker than normal and the supply dewar was ~5 psi (normal supply dewar psi ~3-4psi). So we pulled the transfer line from BUBBA. When we took the transfer line out, there was a partial ice block on the end of the transfer line. This caused the weak plume and the higher than normal supply dewar pressure. I cleared the ice from the end of the transfer line, and we began the fill again. The supply dewar remained steady around 3 psi and the exhaust plume appeared normal. After 25 minutes of filling with a normal pressure, we should have seen the plume change, but of course we didn't. So we pulled the transfer line and measured the amount of helium left in the supply dewar. The supply dewar was empty, which means that we must have added ~25 liters of liquid helium into BUBBA, almost filling it 100%

I have set it up so that anyone can add entries from their own account if they are in the admin group. The trick to this is to force all files to get created as the admin group and users then need to set their umask correctly. First you need to edit your .bashrc and add these (.cshrc people - this is different for you):

  export BLOG_DIR=/Library/WebServer/Documents/blog
  alias nb-add "(umask 002 && nb -a)"

Now source your .bashrc and you should be able to work with the blog:

  source ~/.bashrc

  # List the blog entries
  nb -l

  # Add an entry
  nb-add

To make these changes to the group, I did the following voodoo.

  cd $BLOG_DIR
  chown g+s cache data attachments archive

There may still be some kinks with this! We may experience some turbulence, so keep your seat belts fastened. That mostly worked.

In OSX 10.3, removable media is unmounted when nobody is logged in. To correct this, the following command must be executed.

sudo defaults write \
/Library/Preferences/SystemConfiguration/autodiskmount \
AutomountDiskWithoutUserLogin -bool true

Removable media sources (i.e. FireWire) now stay mounted without being logged in allowing the daily_backup.bash script to run through the crontab

If you want to subscribe to this RSS feed, you might want to get NewNewsWire Lite. The Lite version is free and the link is at the bottom of the page. Click the "Subscribe" button and add this url:

  http://sorcerer.ucsd.edu/blog/rss.xml

#!/bin/bash
## This program backs up the RAWDAT and PDFs on Sorcerer to an
external hard drive FireWire

echo "COPYING RAWDAT TO FIREWIRE"
cp -R ~pmag/RAWDAT /Volumes/FireWire/daily_backup/07
echo "COPYING PDFs TO FIREWIRE"
cp -R ~pmag/Documents/pdfs /Volumes/FireWire/daily_backup/06-pdfs
echo "DONE"

This program is supposed to backup RAWDATA and PDF files daily on Sorcerer to an external drive through a crontab. Currently the external drive is unmounted when nobody is logged in and the script must be run manually. This is not a tar archive, just a full recursive copy.

Use the <pre> blocks to quote things which end up in a gray block. These blocks usually respect your indenting.

   Pmag is fun.
     Squids measure EMU's.

The purpose of this weblog is to catalog and record anything related to the SIO Paleomag lab instruments, computers, measuring techniques, purchases, etc.