File Metadata Schema

From DUNE
Revision as of 14:17, 16 December 2014 by Admin (talk | contribs)
Jump to navigation Jump to search

File Metadata Schema is a description of what excerpted, summary information we store about each file and outside the file itself.

Documents

Drivers

The metadata schema is something that multiple systems rely on

  • detector DAQ systems must ultimtely produce it
  • end-user analyses rely on datasets defined in terms of it
  • production data processing systems need to query it

Currently, the schema to describe data from the the verticle slice test and 35t prototype test run are paramount.

File names

This section describes the naming convention for files. There are two types of files:

collaboration 
raw data files produced by the detectors or by production processing or simulation
personal 
files produced by individual or groups but not in an official production manner.

Collaboration files

Collaboration files are produced in an "official" manner, either via the detectors or through production processing of these files or the producing and processing of simulation. If any file is so produced and their file name conventions are not documented in this section please add them or contact the S&C group.

35t DAQ raw data

The 35t DAQ raw data files are named following this pattern:

35t_r<run number>_s<subrun_number>_<run_mode>_raw.root

For example:

35t_r0000001_s0001_test_raw.root

With:

  • run_number a 7 digit, zero-padded, monotonically increasing integer labeling a "run" of the DAQ where all parameters of the data taking are expected to remain fixed.
  • subrun_number a 4 digit, zero-padded, monotonically increasing integer within one run and labeling a "subrun" of the DAQ. Subruns indicate simple partitions of a run in order to keep file sizes manageable.
  • run_mode is an alphanumeric word describing the selection that the DAQ places on the data (aka, "trigger" but it may not be implemented as a traditional trigger).

Personal Files

Personal files are those produced by an individual or group but not as part of an organized "official" production processing. It is encouraged that the naming conventions of any such files which are shared beyond the individual or group that produced them to be documented in this section. Either document them directly or include links to external documentation. If you know of such files which are not so documented, please do so or contact the S&C group.

SAM

The de'facto choice for a file metadata system is "SAM" from Fermilab. It has a required/default schema. The table in this Redmine page gives a summary. It is up to the experiment to extend and for some elements, interpret, this schema. This extension and interpretation needs to be carefully nailed down and allowed to evolve in a well controlled manner.

File numbers

See below for issues on this topic related to 35t.

SAM supports both subrun and file numbers. It has two fields that pertain on a per-file basis:

runs = [ [<runnumber>, <subrunnumber>, <runtype>], ]
file_partition = <filenumber>

The "runs" variable is a list of 2-tuples or 3-tuples. In the former case, the subrunnumber is dropped. This means that a single file may be recorded as spaning runs and/or subruns. The file_partition is open to interpretation by the experiment (as are the other numbers).

35 ton

This section collects info specific to metadat for 35t raw data file production

File numbering

Some email ca. July 2014 between online and offline people discussed this. The summary:

  • artDAQ writes files to disk and enstore and is the initial source of metadata (Kurt Biery, John Freeman)
  • Monotonically increasing "run number" (likely not starting at zero/one, and tracked/asserted by run control. Subrun and/or file numbers generated by artDAQ (Erik Blaufus)
  • artDAQ will set subrun number and will keep one file per subrun (Kurt Biery)

In terms of SAM, this means either dropping subrunnumber and using the 2-tuple form of runs and using file_partition to count the subrun OR keeping the 3-tuple form and either ignoring file_partition or filling it with a redundant count. My (bv) recomendation is to adopt the latter.

Desire metadata schema

This section collects the metadata schema desired for the 35t, parameter names, types, etc. This will become the fodder for a formal change proposal.

Parameter types:

extenum 
an extensible enumeration. This type takes values from a fixed but extensible set of strings. The set is not allowed to grow unrestricted or via one individuals action. Some consensus among a larger group is required to add to the set.
verstr 
a string left up to the experts who determine the version of something. The values such a type may take should governed by some pattern. See DocDB #9888 section 3.2 for a recommended interpretation of version numbers.
list(type) 
an ordered sequence of elements of type type
N-tuple 
an ordered sequence of exactly N elements of unspecified types.
name type example comment
Basic Parameters (types pre-defined by SAM schema)
file_type extenum raw35t,log indicates category of use, any application that can read files of a given file_type should be able to read all such files. file_type is different than file_format
file_format extenum rawdaq, dk2nu, detsim defines the lowevel format (eg, ASCII, ROOT) and the general schema which the content follows
runs list(3-tuple) [(100,2,"label"),] Following SAM requirements the 3-tuple holds: (run-number, subrun-number, "label"). TBD: specify expectation for label. Note, this is a list; a given file may have multiple run/subrun numbers associated.
...
Schema Extensions (parameters not directly defined by SAM)
file_format_version verstr 1.0, 1.1 See DocDB #9888 section 3.2 for a recommended interpretation of version numbers
...