[Zaphod-Users] IMPORTANT: Upgrades August 13-17
Douglas Larson
djl at artemis.sr.unh.edu
Fri Aug 10 15:48:03 EDT 2007
Zaphod Users may find that jobs are killed and in need of re-submission
next week
because of a software change on the compute nodes. Lustre will be added!
This will
augment scratch space available to users. Lustre is already available as
a directory
called '/scratch' e.g.
djl at h101:~> df -h
Filesystem Size Used Avail Use% Mounted on
/dev/md5 4.6G 1.4G 3.1G 30% /
tmpfs 1.9G 4.0K 1.9G 1% /dev/shm
/dev/md1 99M 65M 30M 69% /boot
/dev/md8 9.2G 3.4G 5.4G 39% /tmp
/dev/md6 46G 9.4G 35G 22% /usr
/dev/md7 9.2G 3.0G 5.9G 34% /var
/dev/md10 2.6T 2.1T 504G 81% /home
s105:/data 2.7T 2.4T 350G 88% /mnt/data07
/dev/loop0 95M 4.7M 86M 6% /lustre/mgs
10.0.0.1 at tcp:/scratch
3.2T 741G 2.4T 24% /scratch
Summary: If you cannot submit jobs, login, or otherwise do zaphod tasks
during the day,
we may be performing some upgrades. We will try and keep the machine in
a running
state during the night.
More information about the Zaphod-Users
mailing list