How to access data on the distributed disk

Setup

You need to checkout from the repository, PhobosUtils, and run a macro in it, to enable you to connected to the nodes on which the disks are attached.

1. cvs checkout PhobosUtils [ENTER]
2. cd PhobosUtils [ENTER]
3. ./enablerootd [ENTER]

This will append information to your .netrc file, to allow you access to the computers on which the disks are sitting.

How to Use it?

In phat, when you want to open a dst, just type
         TPhDST *dst = new TPhDST("oracle:rcf",RUNNUMBER);

And use the TPhDST object as you normally would, i.e. dst->GetNextEvent(), will get the next event. This will loop over all the files with this run number on distributed disk.

How do I find out which files are on the distributed disks?

In the PhobosUtils directory there is a file, lsroot, that lists all the files on distributed disk. So

1. cd PhobosUtils
2. ./lsroot

This will give a listing of all the runs and sequences, and file names that are on distributed disks, and the nodes they are stored on. If you just want to see if a particular run is on the disk, or the online version of the data on the disk for a particular run, type:

1. cd PhobosUtils
2. ./lsroot | grep [RunNumber]

What do I do if the file I want is not on distributed disk?

You should request the files you want from the distributed disk file manager, via email. This person has not yet been assigned.