You will probably find it helpful to familiarise yourself with our intranet and computing web pages. This page is intended mainly for new users to help them get started quickly but others may find it helpful as a reminder.


Many University and Departmental services require that you authenticate yourself using this identifier and we also use it in HEP to control the visibility of parts of our web server. All University members will have been issued with one. It is also used as part of your University e-mail delivery address (<crsid> By default e-mails to this address will be directed to the University e-mail service.

The CRSID is also used for commercially provided services that are accessible via the University single sign-on (GoogleSuite, Microsoft Teams).

HEP e-mail

All HEP communication will normally be directed to your HEP email address therefore it is important that you monitor this e-mail account one way or another.

Your HEP e-mail account is completely separate from the University e-mail service and you should consider whether you want to forward mail from one to the other (or to a third-party email address). Many people find it convenient to do this, rather than trying to keep track of mail on multiple servers. This becomes even more relevant if you get a CERN account. For how to do this and more, see here.

The University Lookup Directory

The University provides a database of all members. You should visit your personal lookup page to update your personal data (Details tab, Edit button) and decide what information you want to provide and what you wish to suppress. We use some of this information to populate our own HEP contact pages.

To conform with the current use of forenames we would ask you to complete the Display name box with your full name as you would wish to be known including any titles.

Under Phone Numbers you will note at least two boxes. Enter your telephone extension number in the first left-hand box. This can be entered in one of the following formats (though remember, Lookup can only generally be seen by University personnel).

01223 337722
+44 1223 337722

If you are proposing to enter more than one telephone number, click the Add button and put the additional number in the next left-hand box. If you are providing more than one telephone number then indicate in the appropriate right-hand box what that number represents (e.g. extension, mobile extension or home). You may wish to distinquish between a Departmental and College extension if you enter both of them.

Use of mobile devices on the network

If you want to use the wired network with your laptop, you will need to provide us with its hardware address so that we can register it. You can use the UniOfCam or eduroam wireless networks if you prefer, but connections to these are treated as being as outside the University, so some services will be restricted. In particular, do not make large data transfers over UniofCam/eduroam. Transfers of more than 10GB in a day are likely to provoke a warning from the University Information Services and a repeat of the offence risks sanctions being imposed.

Also note that any electrical equipment you bring into the group (eg. laptop power units) needs to be electrically safe - otherwise you will be required to remove them from the department. Please consult with Rick Shaw (Room 844) before plugging in your own electrical equipment.

Desktop and disk usage

Your Linux home area (/usera/<username>) is primarly for smaller dynamic files which need a daily backup. It is not intended for large data files which are either static or relatively easy to recreate - these should go on the data disks /rxx (which can be backed-up on request). If in doubt, ask which disk to use.

The Windows home area is mapped onto the Z: disk. It is also visible under Linux as /pchome/<username>. The same comment regarding the use of the Linux home area applies to this area as well.

Some of you may use Ganga for job handling. The default for the Ganga working directory is /usera/<username>/gangadir and this can get very large. You must redirect the gangadir either to the local disk or to an area on /rxx. Similar considerations may apply to other software you might be using.

Large numbers of small files also have a detrimental effect on our backup schedule. Some tools, notably snakemake, can create vast numbers of metadata files in hidden .snakemake directories. Data like these should be directed to the local disk. In the rare cases where these files must be regularly backed up, ensure that they are packaged using a utility like tar. Combine this with the use of data compression to give an even bigger improvement.

Remember that files "deleted" by moving them to a trash folder are not really deleted and remain on disk until the trash folder is emptied. Please ensure that you empty the trash folder regularly.

Please delete any coredump files (core.xxxxx) promptly as these can be very large. You can disable core dumps by adding ulimit -c 0 to your login profile.

NFS may create hidden files (.nfsxxxxxxxxxxxx) that may be very large. These files are created by NFS when a file is deleted but a process still has the file open. Sometimes these are not properly cleaned up automatically. Please remove them.

You may be tempted to install software packages in your home area. Do not do this. The packages can be very large and do not need to be backed up. If you think the package may be generally useful consult with us. Otherwise install to the local disk (/var/work) or the global work area (/work).

Below is a matrix that summarises the recommended usage of the various storage areas that are available.

LocationUsagePersonal softwareThird party softwareDataBacked up
/useraHome areaYesNoNoYes
/var/workLocal workspaceNoYesYesNo
/workGlobal workspaceNoYesNoNo

The Linux desktops are also used to run batch jobs. Please do not shut down any Linux system when you logoff and don't reboot them unless you really have to (and then please check for other use before you do if at all possible).

Large amounts of data traffic sent to the wrong location can cause the system to grind to a halt for everyone. Running a large number of jobs in parallel where the logs write to the same place is one example where problems can occur.

We are happy to advise on computing matters but for questions concerning the software infrastructure of your experiment (LHCb, ATLAS, neutrino) the first port of call should probably be one of the local experiment software experts.

Other information

Finally, you may also find this GoogleDoc helpful.

Steve Wotton
Last update 15 September 2020.