Project

General

Profile

1
Installation:
2
    Check out svn: svn co https://code.nceas.ucsb.edu/code/projects/bien
3
    cd bien/
4
    Install: make install
5
        WARNING: This will delete the current public schema of your VegBIEN DB!
6
    Uninstall: make uninstall
7
        WARNING: This will delete your entire VegBIEN DB!
8
        This includes all archived imports and staging tables.
9

    
10
Maintenance:
11
    VegCore data dictionary:
12
        Regularly, or whenever the VegCore data dictionary page
13
            (https://projects.nceas.ucsb.edu/nceas/projects/bien/wiki/VegCore)
14
            is changed, regenerate mappings/VegCore.csv:
15
            make mappings/VegCore.htm-remake; make mappings/
16
            svn ci -m "mappings/VegCore.csv: Regenerated from wiki"
17
    Important: Whenever you install a system update that affects PostgreSQL or
18
        any of its dependencies, such as libc, you should restart the PostgreSQL
19
        server. Otherwise, you may get strange errors like "the database system
20
        is in recovery mode" which go away upon reimport, or you may not be able
21
        to access the database as the postgres superuser. This applies to both
22
        Linux and Mac OS X.
23

    
24
Data import:
25
    On local machine:
26
        make test by_col=1
27
            See note under Testing below
28
    On vegbiendev:
29
    Ensure there are no local modifications: svn st
30
    svn up
31
    make inputs/upload
32
    For each newly-uploaded datasource: make inputs/<datasrc>/reinstall
33
    Update the auxiliary schemas: make schemas/reinstall
34
        The public schema will be installed separately by the import process
35
    Delete imports before the last so they won't bloat the full DB backup:
36
        make backups/vegbien.<version>.backup/remove
37
    Make sure there is at least 100GB of disk space on /: df -h
38
        The import schema is 75GB, and may use additional space for temp tables
39
        To free up space, remove backups that have been archived on jupiter:
40
            List backups/ to view older backups
41
            Check their MD5 sums using the steps under On jupiter below
42
            Remove these backups
43
    unset version
44
    Start column-based import: . bin/import_all by_col=1
45
        To use row-based import: . bin/import_all
46
        To stop all running imports: . bin/stop_imports
47
        WARNING: Do NOT run import_all in the background, or the jobs it creates
48
            won't be owned by your shell.
49
        Note that import_all will take several hours to import the NCBI backbone
50
            and TNRS names before returning control to the shell.
51
    Wait (overnight) for the import to finish
52
    On local machine: make inputs/download-logs
53
    In PostgreSQL, check that the provider_count and source tables contain
54
        entries for all inputs
55
    tail inputs/{.,}*/*/logs/$version.log.sql
56
    In the output, search for "Command exited with non-zero status"
57
    For inputs that have this, fix the associated bug(s)
58
    If many inputs have errors, discard the current (partial) import:
59
        make schemas/$version/uninstall
60
    Otherwise, continue
61
    make schemas/$version/publish
62
    sudo backups/fix_perms
63
    make backups/upload
64
    On jupiter:
65
        cd /data/dev/aaronmk/bien/backups
66
        For each newly-archived backup:
67
            make <backup>.md5/test
68
            Check that "OK" is printed next to the filename
69
    On nimoy:
70
        cd /home/bien/svn/
71
        svn up
72
        export version=<version>
73
        make backups/analytical_aggregate.$version.csv/download
74
        make backups/analytical_aggregate.$version.csv.md5/test
75
        Check that "OK" is printed next to the filename
76
        In the bien_web DB:
77
            Create the analytical_aggregate_<version> table using its schema
78
                in schemas/vegbien.my.sql
79
        env table=analytical_aggregate_$version bin/publish_analytical_db \
80
            backups/analytical_aggregate.$version.csv
81
    If desired, record the import times in inputs/import.stats.xls:
82
        Open inputs/import.stats.xls
83
        Insert a copy of the leftmost "By column" column group before it
84
        bin/import_date inputs/{.,}*/*/logs/$version.log.sql
85
        Update the import date in the upper-right corner
86
        bin/import_times inputs/{.,}*/*/logs/$version.log.sql
87
        Paste the output over the # Rows/Time columns, making sure that the
88
            row counts match up with the previous import's row counts
89
        If the row counts do not match up, insert or reorder rows as needed
90
            until they do
91
        Commit: svn ci -m "inputs/import.stats.xls: Updated import times"
92
    To remake analytical DB: bin/make_analytical_db &
93
        To view progress:
94
            tail -f inputs/analytical_db/logs/make_analytical_db.log.sql
95

    
96
Backups:
97
    Archived imports:
98
        Back up: make backups/<version>.backup &
99
            Note: To back up the last import, you must archive it first:
100
                make schemas/rotate
101
        Test: make backups/<version>.backup/test &
102
        Restore: make backups/<version>.backup/restore &
103
        Remove: make backups/<version>.backup/remove
104
        Download: make backups/download
105
    TNRS cache:
106
        Back up: make backups/TNRS.backup-remake &
107
        Restore:
108
            yes|make inputs/.TNRS/uninstall
109
            make backups/TNRS.backup/restore &
110
            yes|make schemas/public/reinstall
111
                Must come after TNRS restore to recreate tnrs_input_name view
112
    Full DB:
113
        Back up: make backups/vegbien.<version>.backup &
114
        Test: make backups/vegbien.<version>.backup/test &
115
        Restore: make backups/vegbien.<version>.backup/restore &
116
        Download: make backups/download
117
    Import logs:
118
        Download: make inputs/download-logs
119

    
120
Datasource setup:
121
    Add a new datasource: make inputs/<datasrc>/add
122
        <datasrc> may not contain spaces, and should be abbreviated.
123
        If the datasource is a herbarium, <datasrc> should be the herbarium code
124
            as defined by the Index Herbariorum <http://sweetgum.nybg.org/ih/>
125
    For MySQL inputs (exports and live DB connections):
126
        For .sql exports:
127
            Place the original .sql file in _src/ (*not* in _MySQL/)
128
            Create a database for the MySQL export in phpMyAdmin
129
            mysql -p database <inputs/<datasrc>/_src/export.sql
130
        mkdir inputs/<datasrc>/_MySQL/
131
        cp -p lib/MySQL.{data,schema}.sql.make inputs/<datasrc>/_MySQL/
132
        Edit _MySQL/*.make for the DB connection
133
            For a .sql export, use your local MySQL DB
134
        Install the export according to Install the staging tables below
135
    Add input data for each table present in the datasource:
136
        For .sql exports, you must use the name of the table in the DB export
137
        For CSV files, you can use any name. It's recommended to use a table
138
            name from <https://projects.nceas.ucsb.edu/nceas/projects/bien/wiki/VegCSV#Suggested-table-names>
139
        Note that if this table will be joined together with another table, its
140
            name must end in ".src"
141
        make inputs/<datasrc>/<table>/add
142
            Important: DO NOT just create an empty directory named <table>!
143
                This command also creates necessary subdirs, such as logs/.
144
        If the table is in a .sql export: make inputs/<datasrc>/<table>/install
145
            Otherwise, place the CSV(s) for the table in
146
            inputs/<datasrc>/<table>/ OR place a query joining other tables
147
            together in inputs/<datasrc>/<table>/create.sql
148
        Important: When exporting relational databases to CSVs, you MUST ensure
149
            that embedded quotes are escaped by doubling them, *not* by
150
            preceding them with a "\" as is the default in phpMyAdmin
151
        If there are multiple part files for a table, and the header is repeated
152
            in each part, make sure each header is EXACTLY the same.
153
             (If the headers are not the same, the CSV concatenation script
154
             assumes the part files don't have individual headers and treats the
155
             subsequent headers as data rows.)
156
        Add <table> to inputs/<datasrc>/import_order.txt before other tables
157
            that depend on it
158
    Install the staging tables:
159
        make inputs/<datasrc>/reinstall quiet=1 &
160
        To view progress: tail -f inputs/<datasrc>/<table>/logs/install.log.sql
161
        View the logs: tail -n +1 inputs/<datasrc>/*/logs/install.log.sql
162
            tail provides a header line with the filename
163
            +1 starts at the first line, to show the whole file
164
        For every file with an error 'column "..." specified more than once':
165
            Add a header override file "+header.<ext>" in <table>/:
166
                Note: The leading "+" should sort it before the flat files.
167
                    "_" unfortunately sorts *after* capital letters in ASCII.
168
                Create a text file containing the header line of the flat files
169
                Add an ! at the beginning of the line
170
                    This signals cat_csv that this is a header override.
171
                For empty names, use their 0-based column # (by convention)
172
                For duplicate names, add a distinguishing suffix
173
                For long names that collided, rename them to <= 63 chars long
174
                Do NOT make readability changes in this step; that is what the
175
                    map spreadsheets (below) are for.
176
                Save
177
        If you made any changes, re-run the install command above
178
    Auto-create the map spreadsheets: make inputs/<datasrc>/
179
    Map each table's columns:
180
        In each <table>/ subdir, for each "via map" map.csv:
181
            Open the map in a spreadsheet editor
182
            Open the "core map" /mappings/Veg+-VegBIEN.csv
183
            In each row of the via map, set the right column to a value from the
184
                left column of the core map
185
            Save
186
        Regenerate the derived maps: make inputs/<datasrc>/
187
    Accept the test cases:
188
        make inputs/<datasrc>/test
189
            When prompted to "Accept new test output", enter y and press ENTER
190
            If you instead get errors, do one of the following for each one:
191
            -   If the error was due to a bug, fix it
192
            -   Add a SQL function that filters or transforms the invalid data
193
            -   Make an empty mapping for the columns that produced the error.
194
                Put something in the Comments column of the map spreadsheet to
195
                prevent the automatic mapper from auto-removing the mapping.
196
            When accepting tests, it's helpful to use WinMerge
197
                (see WinMerge setup below for configuration)
198
        make inputs/<datasrc>/test by_col=1
199
            If you get errors this time, this always indicates a bug, usually in
200
                the VegBIEN unique constraints or column-based import itself
201
    Add newly-created files: make inputs/<datasrc>/add
202
    Commit: svn ci -m "Added inputs/<datasrc>/" inputs/<datasrc>/
203
    Update vegbiendev:
204
        On vegbiendev: svn up
205
        On local machine: make inputs/upload
206
        On vegbiendev:
207
            Follow the steps under Install the staging tables above
208
            make inputs/<datasrc>/test
209

    
210
Datasource refreshing:
211
    VegBank:
212
        make inputs/VegBank/vegbank.sql-remake
213
        make inputs/VegBank/reinstall quiet=1 &
214

    
215
Schema changes:
216
    Remember to update the following files with any renamings:
217
        schemas/filter_ERD.csv
218
        mappings/VegCore-VegBIEN.csv
219
        mappings/verify.*.sql
220
    Regenerate schema from installed DB: make schemas/remake
221
    Reinstall DB from schema: make schemas/public/reinstall schemas/reinstall
222
        WARNING: This will delete the current public schema of your VegBIEN DB!
223
    Reinstall staging tables: . bin/reinstall_all
224
    Sync ERD with vegbien.sql schema:
225
        Run make schemas/vegbien.my.sql
226
        Open schemas/vegbien.ERD.mwb in MySQLWorkbench
227
        Go to File > Export > Synchronize With SQL CREATE Script...
228
        For Input File, select schemas/vegbien.my.sql
229
        Click Continue
230
        In the changes list, select each table with an arrow next to it
231
        Click Update Model
232
        Click Continue
233
        Note: The generated SQL script will be empty because we are syncing in
234
            the opposite direction
235
        Click Execute
236
        Reposition any lines that have been reset
237
        Add any new tables by dragging them from the Catalog in the left sidebar
238
            to the diagram
239
        Remove any deleted tables by right-clicking the table's diagram element,
240
            selecting Delete '<table name>', and clicking Delete
241
        Save
242
        If desired, update the graphical ERD exports (see below)
243
    Update graphical ERD exports:
244
        Go to File > Export > Export as PNG...
245
        Select schemas/vegbien.ERD.png and click Save
246
        Go to File > Export > Export as SVG...
247
        Select schemas/vegbien.ERD.svg and click Save
248
        Go to File > Export > Export as Single Page PDF...
249
        Select schemas/vegbien.ERD.1_pg.pdf and click Save
250
        Go to File > Print...
251
        In the lower left corner, click PDF > Save as PDF...
252
        Set the Title and Author to ""
253
        Select schemas/vegbien.ERD.pdf and click Save
254
        Commit: svn ci -m "schemas/vegbien.ERD.mwb: Regenerated exports"
255
    Refactoring tips:
256
        To rename a table:
257
            In vegbien.sql, do the following:
258
                Replace regexp (?<=_|\b)<old>(?=_|\b) with <new>
259
                    This is necessary because the table name is *everywhere*
260
                Search for <new>
261
                Manually change back any replacements inside comments
262
        To rename a column:
263
            Rename the column: ALTER TABLE <table> RENAME <old> TO <new>;
264
            Recreate any foreign key for the column, removing CONSTRAINT <name>
265
                This resets the foreign key name using the new column name
266
    Creating a poster of the ERD:
267
        Determine the poster size:
268
            Measure the line height (from the bottom of one line to the bottom
269
                of another): 16.3cm/24 lines = 0.679cm
270
            Measure the height of the ERD: 35.4cm*2 = 70.8cm
271
            Zoom in as far as possible
272
            Measure the height of a capital letter: 3.5mm
273
            Measure the line height: 8.5mm
274
            Calculate the text's fraction of the line height: 3.5mm/8.5mm = 0.41
275
            Calculate the text height: 0.679cm*0.41 = 0.28cm
276
            Calculate the text height's fraction of the ERD height:
277
                0.28cm/70.8cm = 0.0040
278
            Measure the text height on the *VegBank* ERD poster: 5.5mm = 0.55cm
279
            Calculate the VegBIEN poster height to make the text the same size:
280
                0.55cm/0.0040 = 137.5cm H; *1in/2.54cm = 54.1in H
281
            The ERD aspect ratio is 11 in W x (2*8.5in H) = 11x17 portrait
282
            Calculate the VegBIEN poster width: 54.1in H*11W/17H = 35.0in W
283
            The minimum VegBIEN poster size is 35x54in portrait
284
        Determine the cost:
285
            The FedEx Kinkos near NCEAS (1030 State St, Santa Barbara, CA 93101)
286
                charges the following for posters:
287
                base: $7.25/sq ft
288
                lamination: $3/sq ft
289
                mounting on a board: $8/sq ft
290

    
291
Testing:
292
    Mapping process: make test
293
        Including column-based import: make test by_col=1
294
            If the row-based and column-based imports produce different inserted
295
            row counts, this usually means that a table is underconstrained
296
            (the unique indexes don't cover all possible rows).
297
            This can occur if you didn't use COALESCE(field, null_value) around
298
            a nullable field in a unique index. See sql_gen.null_sentinels for
299
            the appropriate null value to use.
300
    Map spreadsheet generation: make remake
301
    Missing mappings: make missing_mappings
302
    Everything (for most complete coverage): make test-all
303

    
304
WinMerge setup:
305
    Install WinMerge from <http://winmerge.org/>
306
    Open WinMerge
307
    Go to Edit > Options and click Compare in the left sidebar
308
    Enable "Moved block detection", as described at
309
        <http://manual.winmerge.org/Configuration.html#d0e5892>.
310
    Set Whitespace to Ignore change, as described at
311
        <http://manual.winmerge.org/Configuration.html#d0e5758>.
312

    
313
Documentation:
314
    To generate a Redmine-formatted list of steps for column-based import:
315
        make inputs/ACAD/Specimen/logs/steps.by_col.log.sql
316
    To import and scrub just the test taxonomic names:
317
        inputs/test_taxonomic_names/test_scrub
318

    
319
General:
320
    To see a program's description, read its top-of-file comment
321
    To see a program's usage, run it without arguments
322
    To remake a directory: make <dir>/remake
323
    To remake a file: make <file>-remake
(2-2/5)