FAQ: Data, Analysis & Publications
How do I access the FACET-II Servers?
If you would like to get access to FACET-II servers, for example to pull
up the facethome control panel, you need to do the following...
Part 1: (this is described in more detail here)
1). Get a SLAC unix account.
2). Generate a public rsa key on your SLAC unix account.
2). Email the control-system-admin copying Christine Clarke (firstname.lastname@example.org) to request permission to access the FACET-II servers as a physics user, and include your SLAC unix account name.
Once you've done this, Ken will grab your RSA key from your unix account and grant you access permission to ssh into the FACET server.
To access the FACET server and pull up the facethome control panel, you
need to do the following...
1) ssh to a SLAC unix machine (e.g. rhel6-32.slac.stanford.edu, rhel6-64.slac.stanford.edu). Use the -Y option to allow X11 forwarding. Or use fastx.slac.stanford.edu
2) ssh to the mcclogin machine. If you have a SLAC unix account, you can access this machine with the same user name and password.
3) ssh to facet-srv01 with the user name fphysics. This does not require a password. After connecting, you will be asked to type in a number corresponding to your unix username. This will load your own personal user settings and place you in your home directory. It's a little weird, but just roll with it.
Note: You can also log into facet-srv20 which is a server local to sector 20 (rather than MCC).
4) Type 'facethome' to bring up the FACET control panel. DO NOT MESS WITH THINGS THAT YOU DO NOT UNDERSTAND! One wrong click of a button could cause a world of grief to your fellow experimentalists.
How do I get data to [anywhere else]?
Data is saved to /u1/facet/matlab/data if taken with a matlab GUI like profile monitor or TCAV or Correlation Plot.
This is only accessible from within the accelerator controls network- by necessity, a rather well protected network.
Obviously you want to get data to the outside world.
There are quite a few ways of doing this... scp from facet-srv01 to personal space on mcclogin works but you'll run into the issue of limited disk space pretty fast.
The pwfa wiki has a solution. You need to be on the internal SLAC network to view the pwfa wiki (VPN in).
The files are also accessible on a web interface:
If you are saving other things like archive plots or images, you may want to create a directory for yourself in /u1/facet/physics
Copy files here and access through the web.
If you are writing your own DAQ and want to save it somewhere convenient, /u1 is not the place- especially if the files are large! See the other FAQ item about where to save data. If you save data to the place recommended in this FAQ, data is automatically mirrored to a location within SLAC but outside of the controls network. Then you simply scp to your home institution.
Mirrored data is available here:
You can ssh to rhel6-64.slac.stanford.edu or noric.slac.stanford.edu or flora.slac.stanford.edu from outside of SLAC and copy the data. If you are inside of SLAC, or using VPN, we recommend going through lcls-prod02 .
scp email@example.com :/nfs/slac/g/facetdata/nas/nas-li20-pm00/E200/2014/<date_dir>/<data_dir> <dir on local machine>
For faster and more robust copying, use rsync:
rsync -av firstname.lastname@example.org :/nfs/slac/g/facetdata/nas/nas-li20-pm00/E200/2014/<date_dir>/<data_dir> <dir on local machine>
I've saved data through a matlab GUI... where is it?
Any data taken using the MATLAB GUIs found in facethome (dispersion scans, wire scans, TCAV scans etc) is saved and backed up on the controls network NFS. If you are using one of the facet-srv machines you can find the data here:
and if you are away from SLAC, the data is conveniently mirrored here:
This site can only be accessed if you are on the VPN.
Data saved to the NAS devices in Sector 20 are backed up weekly to:
This data can be accessed by ssh or scp on any SLAC SCCS machine (for instance rhel6-32.slac.stanford.edu, rhel6-64.slac.stanford.edu, or mcclogin).
Where can I save my data to?
We have 20 TB of storage local to sector 20. The drives are networked and connected to our devices through a single switch for maximum data transfer speeds.
Log into facet-srv20. You can launch a terminal to facet-srv20 from facethome if you are on one of the sunray computers (OPIs).
The three networked drives are /nas.
You can save here. Create a directory for your experiment.
The drives get backed up daily to our mirror:
It has the same path structure as what you would see on facet-srv20 with /nas as the top directory.
This is accessible from anywhere in slac- you don't need to be on the controls network. If you are off-site, this is just one ssh connection away.
Use the mirror to access your data.
If you transfer your data from the nas drives, you will cause unwelcome network traffic that could impede data taking.
Also, you only have read access for the mirrored data making it safer (an errant keystroke can't do much damage!).
Of course, there is a time delay as the back up only runs once. If this is not okay for you, you can contact Christine to transfer data earlier.
You can also use matlab on facet-srv20 to analyse data locally or even on the fly. Just run fmatlab from a terminal on facet-srv20.
Why can't I see facet-data?
A couple of people have said that the directory goes missing.
Please report this to Christine if this happens but note that it may just be that you haven't requested that the drive be mounted...
The NFS automounter does not mount the facetdata directory until it is requested.
It is requested by doing a "cd" into the directory, or starting I/O on a file in there.
There are lots and lots of shared areas under "/nfs/slac/g/" -- these are only mounted on request. And then unmounted automatically when they are idle. That's the whole point of the automounter.
You can use the "amdmap" command to see what is available (you can't see what is available by doing an 'ls' on "/nfs/slac/g/".
Are there SLAC/AD requirements for publications derived from work done at FACET and Test Beam Facilities?
There are three circumstances to consider:
A. Publication generated using SLAC resources (e.g. SLAC beam, services, people ..). This work must get a SLAC publication number and be submitted to the SLAC document repository, SciDoc. THIS IS A REQUIREMENT OF THE SLAC DOE CONTRACT.
B. Publication has a SLAC co-author. It is the responsibility of the principal author to ensure that SLAC, SLAC personnel and the SLAC contract is appropriately acknowledged. Submitting the the publication to SciDoc will help ensure that this is done correctly.
C. Publication has a SLAC (primary) author. In this case the AD Publication Policy and Guidelines apply. Click on the link to get the policy document.
If you fulfil any of these circumstances, please read the Publication Policy below.
2). Date - April 10, 2012
3). For - publications in peer-reviewed journals, conferences, research repositories and formal Laboratory scientific reports.
4). Applies - to all Accelerator Directorate (AD) personnel listed as authors or co-authors in publications. Also to FACET & Test Facility users who have used SLAC resources to generate the publication.
Credit - all publications must acknowledge funding source (e.g. Work supported by Department of Energy contract DE-AC02-76SF00515).
Author affilations - must be shown on title page
SciDoc - All publications must have a SLAC publication number. It is the lead author’s responsibility to request a publication number and upload the final draft of the publication to the SciDoc repository.
Review & Approval - All publications where the lead author is in AD must be reviewed and approved by their division before being submitted. This isn't required for publications where the lead author is not in AD.
- Coauthors - must make a substantive contribution and have the right to read and comment on the publication before submission.
- Collaboration - When collaboration between groups or institutes is involved, special care should be taken to ensure proper credit is given for the contributions of each.
- Acknowledgement - see policy document for when to acknowledge contributions from non authors.
- References - referencing should gives credit to those whose research has informed or led to the work in question
SciDOC Public Site: https://www-public.slac.stanford.edu/scidoc/
SLAC SciDOC: https://www-internal.slac.stanford.edu/scidoc/
Faculty Handbook: http://www.stanford.edu/dept/provost/faculty/policies/handbook/
Research Policy Handbook: http://rph.stanford.edu
(d). SLAC Policy and Procedure Library: https://slacspace.slac.stanford.edu/sites/ipm/policies/Lists/Policy%20Repository/AllItems.aspx
(e). American Physical Society statement: http://www.aps.org/policy/statements/02_2.cfm
(f). IEEE Policy on Authorship: http://www.ieee.org/documents/opsmanual.pdf http://www.ieee.org/publications_standards/publications/rights/publishing_principles.html
(g). Stanford OTL invention disclosure form: https://otlportal.stanford.edu/wd/
How do I register a publication?
Please register documents through SciDoc. Please report problems with the website to Christine Clarke (email@example.com) who can pass on the comments to the appropriate person.