Storage on the cluster

[Gregory Houchins]

Hi all,

Many of you may have realized that all of the jobs were canceled late last
night due to the storage on the cluster being completely full. I would like
to bring to everyone's attention that there is limited storage on the
cluster and all users should be cognizant of that. When users forget this
fact, we all experience an interruption in service as was seen last night.

For you reference you can check the state of the memory by typing

df -h

/dev/mapper/raid_vg-home_lv is the relevant memory location that /home/ is
mounted to.


You can also check your usage by running

du -sh /home/*username*

This will give you a summary (-s) in human readable units (-h) of how much
memory you are taking up.

There is no set number for how much memory each user can use but those
users using around 500GB or more, are using a disproportionally high amount
of storage and should strongly consider removing unimportant data.

Thanks!

Gregory Houchins | WH3402 | 412-268-7482
Arjuna System Administrator
PhD Candidate, Physics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.andrew.cmu.edu/mailman/private/arjuna-users/attachments/20180130/fd138627/attachment.html>