Newer
Older
# Getting Started
## Introduction
The APS Data Management System is a system for gathering together experimental data,
metadata about the experiment and providing users access to the data based on a users
role. This guide is intended to provide beamline users an introduction to the basic
use of the Data Management System at the beamline. This process will involve creating an
experiment, associating users with this experiment and then adding data, in the form of
files, to the experiment.
### Setting up the Environment
On beamline Linux computers, users can set up the environment for using the Data
Management System by executing a setup script which is created as the Data Management
software is installed. This script is executed as
> /home/DM\_INSTALL\_DIR/etc/dm.setup.sh
where DM\_INSTALL\_DIR is the deployment directory for this beamline. This script
will set up a number of environment variables which define items such
as URLs for the various data services, the station name in the DM, the location
of a file which defines the login to be used when running commands and adding
the path to DM system commands to the PATH.
## DM System Commands
-----
After execution of the setup script, the PATH will include the path to commands
for interacting with the DM System. A list of these commands, at the time of
this writing is shown below.
These commands follow some conventions, like adding a --help for convenience. When
the environment variable DM\_LOGIN\_FILE is defined and points to a file that
contains a username/password pair (in the form 'username|password') this information
is used for authentication when executing the commands. In practice the account
defined here should have the role of station manager for the beamline.
### DM Command usage examples
#### Getting some data into the system
As stated early, the experiment is the central object for creating entries in
the Data Management System. For convenience there are two commands which can
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
provide examples of creating experiments, assigning users & other attributes and
loading data into the system. These commands are
> dm-STATION\_NAME-upload
>
> dm-STATION\_NAME-daq
Where STATION\_NAME is the station name, in lower case with '-' removed. The two
commands here are differentiated by the two methods for moving data into the
system _upload_ and _daq_. moving files using _upload_ will move all files present
in a directory at the time the command is used while _daq_ moves will monitor
files while active and move new files as well. Both of these commands require
two parameters --experiment and --data-directory. With only these required parameters,
these commands will create an experiment named by --experiment and will move files
from --data-directory. Other parameters here will allow for specifying items such
as a list of users, specifying the destination directory (relative to the storage
directory), process a workflow using the data, etc.
Without the optional parameters, use of these commands would look like
> dm\-STATION_NAME\-upload --experiment=exp1 --data-directory=/home/bluser/data/2010-03
This command will create an experiment with no users and move files from
/home/bluser/data/2010-03 to STORAGE\_DIR/STATION\_NAME/EXP\_NAME on the data storage
server where
- STORAGE\_DIR is the base storage location on the storage sever
- STATION\_NAME is the station name defined on the experiment server
- EXP\_NAME is the experiment name defined by --experiment, _exp1_ in this case
Adding other parameters to this command can add more information to the experiment
such as --users to add a list of users to the system. Other parameters such as
--dest-directory and --root-path allow customization of the path to the file on
the data storage server. Some beamlines for instance have relied on particular
directory structure for legacy analysis code.
Sector 8-ID-I for instance has a file structure ROOT\_DIR/RUN/usernameYYYYMM with
- ROOT\_DIR base level for experiment data. In DM this is STORAGE\_ROOT/STATION\_NAME
- RUN is the APS Run cycle such as 2019-1 (year and cycle 1, 2 or 3)
- username is some form to identify the experimenter
- YYYY is the four digit year.
- MM is a 2 digit month, i.e. 02 for February
an example of this would be STORAGE\_ROOT/8idi/2017-2/dufresne201707. Here we
would use a command like
> dm-8idi-upload --experiment=dufresne201707
> --data-directory=/net/s8iddata/export/8-id-i/2017-2/dufresne201707
Sectors 33 & 34 use a file structure under ROOT\_DIR/username/YYYYMMDD/EXP_NAME/.
Here
- ROOT\_DIR base level for experiment data. In DM storage server this is
STORAGE\_ROOT/STATION_NAME
- YYYY four digit year
- MM 2 digit month
- DD 2 digit day
an example of this is ROOT\_DIR/jenia/20140820/FeCrbilayers. Here we would
use a command like
> dm-33bm-upload --experiment=FeCrbilayers
> --data-directory=/net/s33data/export/33bm/data/jenia/20140820/FeCrbilayers
#### Creating experiments based on the ESAF/Proposal
For convenience, at APS, it is possible to make use of the Proposal and ESAF systems
to create experiments based off the existing entries in these systems to populate
the list of users in the system. the --proposal-id and --esaf-id options on commands
like dm-STATION\_NAME-upload and dm-create-experiment will add a list of users which
are defined in the esaf or proposal to the created experiment.
#### Displaying information in the system
Other DM commands show useful information, and can adjust what and how the
information is displayed. `dm-list-experiments` lists the experiments showing
info on the experiment name, experiment type, station, and start date. adding the
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
option --display-keys can allow the user to display more or less information.
--display-keys=name will display only the experiment name, and --display-keys=ALL
will display all keys associated with the experiment (one experiment line is show
as example).
```
startDate=2019-12-09 18:35:26.738217-06:00 name=yuyin201912 description=Study of dynamics and aging in solvent segregation driven colloidal gel in a binary solvent (Proposal id: 64326) rootPath=2019-3 experimentStationId=5 id=3579 experimentTypeId=9 experimentStation={u'description': u'Sector 8 ID I', u'id': 5, u'name': u'8IDI'} experimentType={u'id': 9, u'name': u'XPCS8', u'description': u'XPCS Group (Sector 8)'}
```
Adding the --display-format will change how the data is displayed. By default
the output is a simple dictionary, --display-format=html will give HTML output.
```
<tr> <td>2019-12-09 18:35:26.738217-06:00</td> <td>yuyin201912</td> <td>Study of dynamics and aging in solvent segregation driven colloidal gel in a binary solvent (Proposal id: 64326)</td> <td>2019-3</td> <td>5</td> <td>3579</td> <td>9</td> <td>{u'description': u'Sector 8 ID I', u'id': 5, u'name': u'8IDI'}</td> <td>{u'id': 9, u'name': u'XPCS8', u'description': u'XPCS Group (Sector 8)'}</td> </tr>
```
Specifying --display-format=pprint will give a nicer (prettier) output,
breaking nicely on different lines, indenting properly where some items are objects.
```
{ u'description': u'Study of dynamics and aging in solvent segregation driven colloidal gel in a binary solvent (Proposal id: 64326)',
u'experimentStation': { u'description': u'Sector 8 ID I',
u'id': 5,
u'name': u'8IDI'},
u'experimentStationId': 5,
u'experimentType': { u'description': u'XPCS Group (Sector 8)',
u'id': 9,
u'name': u'XPCS8'},
u'experimentTypeId': 9,
u'id': 3579,
u'name': u'yuyin201912',
u'rootPath': u'2019-3',
u'startDate': u'2019-12-09 18:35:26.738217-06:00'}
```
#### Creating experiments based on the ESAF/Proposal
For convenience, at APS, it is possible to make use of the Proposal and ESAF systems
to create experiments based off the existing entries in these systems to populate
the list of users in the system. the --proposal-id and --esaf-id options on commands
like dm-STATION\_NAME-upload and dm-create-experiment will add a list of users which
are defined in the esaf or proposal to the created experiment.
### dm-station-gui
#### Overview
One of the methods to manage processes in the APS Data Management System is the
application `dm-station-gui`. This application is a PyQt application which gives
access to add/modify/control items such as experiments, file transfer to storage
location ('daq' and 'uploads'), workflows and processing jobs. An example of
this application is shown in the figure below.
#### Experiments
`dm-station-gui` opens showing a tab that lists experiments which have been added
to the DM System. At the beamline, these experiments generally will correspond
to an accepted proposal in the APS Proposal system. Experiments in the DM
System define an entity that ties sets of managed data together. When an
experiment is selected by double clicking or by clicking and then clicking the
*Use Selected* button, the contents of the Experiment tab changes to give
details about that experiment. Much of this information is pulled from the
proposal/ESAF databases. This is shown in the image below.
This view of the data enable a couple of key features of the DM System. The
ability to associate files with the Management system, and the ability to set
which users have access to the system and therefore which data they can access.
Selecting users that will have access to experiment data is done by clicking the
*Modify Users* button below the user list and then selecting users in the right
list & pressing the arrow button between the lists to add users or selecting
users from the left list and clicking the arrow button to delete users from the
list. Click the _Save_ button at the bottom to accept the changes and go back
to the Experiment detail or click *Back* button in the upper left to exit back
to Experiment detail without saving. This view is shown below.
#### Getting Files into Data Management
The overall purpose of this system is to get data into the system and providing
access control to that data. This means linking the data files to the users in
an experiment. For each beam station there is storage location defined for that
station. On the `Experiments` tab there are a couple of items relevant to
getting the data onto that storage location. These items are shown in the image
below. Files transferred onto the storage location will go into:
*STORAGE_LOCATION/(storage_root_path)/(experiment_name)*. Files to go into this
directory are specified in the entry "Data Directory or Single File Path". Note
that if the data directory is specified, only the files will go into the
storage location, not a new sub-directory with the transferred files. Any
sub-directories will be copied as will the contents.
Once the the storage location and source are defined the transfer can be started
in one of two ways:
- A monitor can be placed on the directory and any new files/sub-directories in the directory will be transferred if you select `Start DAQ`
- A one time copy will happen and all files/sub-directories currently in the directory will be copied if you select `Start Upload`
When a transfer is started, the display will switch to either the `DAQs` or
`Uploads` tab. The tab will show the status of all transfers that have been
started since the last restart of the system. Status of a particular transfer is
shown by color
- green: Done success
- yellow: running
- red: Done with an errors
An example of this is shown below.
Clicking on a particular transfer, on `DAQs` or `Uploads` will switch the
view to show detail of that transfer including errors.
#### Workflows and processing data
The tab 'Workflows' allows defining a workflows which will provide a template for processing data. A workflow is simply a set of defined steps that will process data. Workflows each step in the workflow executes a command on the host computer. Commands can
* be as simple as a single command such as listing a directory
* can transfer files between computers
* can launch & monitor jobs on a remote cluster
* can do just about anything that can be scripted on a computer.
Workflow steps allow defining inputs that are defined at runtime and can also create outputs that can be used in following steps.
The `Workflows` tab is shown below and contains a list of workflows that have been defined.
Clicking on a workflow and then selecting the 'Inspect' button you will be able to examine and possibly modify the steps in the workflow.