ADMIRAL 20100304 meeting TG

Meeting with TG, silk group

When: 04-Mar-2010, 10:00

Where: B17

Present: TG, DG, GK

The goal is to understand better some initial requirements of a specific researcher (TG)

(The agenda is kept simple so reassure researchers that frequent meetings will not be unduly time consuming.)

Data volumes and types; frequency of update/access? - use questionnaire returns as starting point.
Ask if it is easy to provide samples of data (for testing and public demonstration)
Is delayed access a problem (e.g. 2 days notice)?
Directory structure conventions?
Who manages any current shared network drive, esp. access control (if applicable).
Use for data file versioning?
IE users: which version.
Saving to LSDS - 3 choices - direct use, copy from local, auto harvest from local.

Data volumes and types; frequency of update/access?

Total volume in 30-100Gb range
raw data files are small
processed data files are larger
raw data is not updated after creation
processed data is frequently (re-)accessed
raw data is required occasionally for reprocessing, prompt access is not critical (but note this is a minority by volume)
indications are that all data will be kept comfortably on fast SAS storage

Is it easy to provide samples of data

Directory structure conventions?

historically, the research group data has used such conventions
TG's own data does not, depending rather on internal content
- (data in spreadsheet / database tends to be quite self-describing)

Who manages network drive, esp. access control (if applicable).

Versioning of data files useful?

probably not
- raw data is never changed
- processed data may be regenerated, but not manually modified
uses home-grown software, with ad-hoc code versioning by filename

Saving to LSDS - 3 choices - direct use, copy from local, auto harvest from local.

more comfortable with local copy
likes auto-harvesting approach, maybe with option to delete on copy
- (must be contained within university network - can't use DropBox.com)
currently, data isn't cleaned up and local computer gets overloaded
- sometimes leads to local system crashing
- options to clean up local data after copy to LSDS are attractive