-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathchanges.html
7171 lines (6686 loc) · 374 KB
/
changes.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
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
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
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
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<html lang="en-US">
<head>
<title>ERDDAP™ - Changes</title>
<meta charset="UTF-8">
<link rel="shortcut icon" href="https://coastwatch.pfeg.noaa.gov/erddap/images/favicon.ico">
<link href="../images/erddap2.css" rel="stylesheet" type="text/css">
<meta name="viewport" content="width=device-width, initial-scale=1">
</head>
<body>
<table class="compact nowrap" style="width:100%; background-color:#128CB5;">
<tr>
<td style="text-align:center; width:80px;"><a rel="bookmark"
href="https://www.noaa.gov/"><img
title="National Oceanic and Atmospheric Administration"
src="../images/noaab.png" alt="NOAA"
style="vertical-align:middle;"></a></td>
<td style="text-align:left; font-size:x-large; color:#FFFFFF; ">
<strong>ERDDAP</strong>
<br><small><small><small>Easier access to scientific data</small></small></small>
</td>
<td style="text-align:right; font-size:small;">
<br>Brought to you by
<a title="National Oceanic and Atmospheric Administration" rel="bookmark"
href="https://www.noaa.gov">NOAA</a>
<a title="National Marine Fisheries Service" rel="bookmark"
href="https://www.fisheries.noaa.gov">NMFS</a>
<a title="Southwest Fisheries Science Center" rel="bookmark"
href="https://www.fisheries.noaa.gov/about/southwest-fisheries-science-center">SWFSC</a>
<a title="Environmental Research Division" rel="bookmark"
href="https://www.fisheries.noaa.gov/about/environmental-research-division-southwest-fisheries-science-center">ERD</a>
</td>
</tr>
</table>
<div class="standard_width">
<h1 style="text-align:center;">ERDDAP™ Changes</h1>
<p>ERDDAP™ is a great example of
<a rel="bookmark" href="https://en.wikipedia.org/wiki/User_innovation">User-Driven Innovation<img
src="../images/external.png" alt=" (external link)"
title="This link to an external website does not constitute an endorsement."></a>,
where product innovation often comes from consumers (ERDDAP™ users), not just the producers (ERDDAP™ developers).
Over the years, most of the ideas for new features and changes in ERDDAP™ have come from users.
Those users are credited below for their great ideas. Thank you! Please keep
those great suggestions coming!
<p>Here are the changes associated with each ERDDAP™ release.
<!--
<li> FLAWED! Announce after fixed.
New data file type: Users can now request gridded and tabular data
in the Igor .itx format. (Flawed intro in v1.74.)
<li> In .nc files written by ERDDAP™,
char variables to be interpreted as individual chars have the attribute
"charset"="ISO-8859-1".
In .nc files read by ERDDAP™, char variables with ... will be interpreted
with the specified charset.
-->
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.25" href="#changes2.25" rel="bookmark">Changes</a>
in ERDDAP™ version 2.25 (released 2024-10-31)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>EDDTableFromFiles can now support queries with only derived outputs (globals, jexl script, or variables).
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>Version 2.25 requires Java 21 or newer. This is the LTS version and has been available for over a year.
<br>
<li>The SharedWatchService is now the default. If you need to disable it, please contact chris.john at noaa.gov to let me know,
so I can improve it in future versions and add:
<br><kbd>
<useSharedWatchService>false</useSharedWatchService>
</kbd> to your setup.xml.
<br>
<li>The ERDDAP™ servlet will now start at server startup. Which means datasets will begin loading immediately instead of
waiting until a request is made.
<br>
<li>The removeMVRows parameter in EDDTableFromMultidimNcFiles will now have an effect. Setting it to false may
significantly speed up some queries, but this may not be suitable for all datasets. For more information see
the <a href="https://erddap.github.io/setupDatasetsXml.html#removeMVRows">description of the parameter</a>.
<br>
<li>Datasets (EDDTableFromNcFiles and EDDGridFromNcFiles) using zarr files are now supported. They must include "zarr"
in either the fileNameRegex or pathRegex. See the <a href="https://erddap.github.io/setupDatasetsXml.html#zarr">zarr
secion in the datasets documentation</a> for more details.
<br>
<li>New dataset type, EDDTableFromParquetFiles is now supported. See the
<a href="https://erddap.github.io/setupDatasetsXml.html#EDDTableFromParquetFiles">EDDTableFromParquetFiles
secion in the datasets documentation</a> for more details.
<br>
<li><a href="https://prometheus.io/">Prometheus metrics</a> are now available at <kbd>/erddap/metrics</kbd>.
<br>
<li>A new XML parser implementation is available. This new parser allows using XInclude in datasets.xml.
Thanks to Ayush Singh for the feature.
<br>
<li>New parameter in datasets.xml to control unusual activity emails. unusualActivityFailPercent defaults to the old value of 25%.
Thanks to Ayush Singh for the feature.
<br>
<li>New parameter in setup.xml that controls if dataset loading errors are shown on the status.html page. It defaults to true, to
disable dataset errors on the status page, set showLoadErrorsOnStatusPage to false:
<kbd>
<showLoadErrorsOnStatusPage>false</showLoadErrorsOnStatusPage>
</kbd>
<br>
<li>Some small changes and bug fixes.
<br>
</ul>
<li><strong>For ERDDAP™ Developers:</strong>
<ul>
<li>Testing separated to unit and integration (slow) tests. Also more tests enabled and tests have been made less flaky.
<br>
<li>Error Prone (some checks still disabled) and Spot Bugs integrated through Maven.
<br>
<li>Full code base formatted to match the Google Style Guide.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.24" href="#changes2.24" rel="bookmark">Changes</a>
in ERDDAP™ version 2.24 (released 2024-06-07)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>New color palette EK80 for acoustic datasets available.
Thanks to Rob Cermak for this.
<br>
<li>Fixen an issue where EDDTableAggregateRows did not show proper ranges from all children.
Thanks to Marco Alba for the fix and bug report.
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: SECURITY CHANGE: Google Authentication might require changes to your CSP.
<p>Specifically, you may also need to add <kbd>https://accounts.google.com/gsi/style</kbd>
to <kbd>stlye-src</kbd> and <kbd>https://accounts.google.com/gsi/</kbd> to <kbd>connect-src</kbd>.
For the <kbd>script-src</kbd> you can now use <kbd>https://accounts.google.com/gsi/client</kbd>.
<p>For more information you can go to the <a rel="help"
href="https://developers.google.com/identity/gsi/web/guides/get-google-api-clientid#content_security_policy">Google
page</a>
about CSP configuration.
<br>
<li>New Shared Watch Service. This is a new option for watching directories for updates. It has one thread for
each filesystem instead of one thread per dataset. Most likely this will drastically reduce the number of
threads used to watch for changes. It does mean all datasets get updated together instead of each dataset
having its own update frequency. Most likely this will mean more frequent updates for most datasets.
<br><p>To enable this add <kbd>
<useSharedWatchService>true</useSharedWatchService>
</kbd> to your setup.xml.</p>
<br>Please do try this and report back how it works for you to chris.john at noaa.gov.
<br>
<li>Fix for incorrect var names in logs.
Thanks to Ayush Singh for the fix.
<br>
<li>Some small changes and bug fixes.
<br>
</ul>
<li><strong>Improvements for ERDDAP™ developers:</strong>
<ul>
<li>Support for local development using Docker.
Thanks Matt Hopson and Roje.
<br>
<li>Support for local development using Jetty and documentation improvements.
Thanks Micah Wengren.
<br>
<li>Changes to tests to reduce issues cross platform.
Thanks Shane St. Savage.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.23" href="#changes2.23" rel="bookmark">Changes</a>
in ERDDAP™ version 2.23 (released 2023-02-27)</h2>
Note that this release was done by Bob Simons, thereby showing that he
is still around and active during the transition to Chris John, his successor.
Stating with this release, all code changes are being done by Chis John, unless specified otherwise.
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>(None)
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: SECURITY CHANGE: Google Authentication is now accomplished via the new
Google Identity Services library which is part of "Sign In with Google".
Google's support for the old "Google Sign In" system will be discontinued 2023-03-31.
So if you use Google Authentication in your ERDDAP™ installation,
you MUST update to ERDDAP™ v2.23+ before then.
(Bob is sorry for the short notice. It's Bob's fault.)
<br>
<li>IMPROVED: NCCSV is now v1.2. The change is that the files are now UTF-8-encoded files (they were ASCII)
and so can now include any Unicode character as is, without encoding as \u<i>hhhh</i>, although that is still allowed.
<br>When writing NCCSV files, ERDDAP™ now writes v1.2 files.
<br>ERDDAP™ will still read NCCSV files that follow the v1.0 and v1.1 specification.
<br>Thanks to Pauline-Chauvet, n-a-t-e, and thogar-computer for suggesting this
and doing the tests to ensure various spreadsheet programs can import UTF-8 files.
Thanks to Bob Simons for this code change.
<br>
<li>NEW: The status.html web page now has a line near the top which indicates
which dataset loadDatasets is currently loading and related statistics,
or none if no dataset is being loaded.
This can be very helpful to ERDDAP™ administrators trying to figure out why
loadDatasets is taking so long.
Also, the nGridDatasets, nTableDatasets, and nTotalDatasets counts below that
are now instantaneous (previously, they were as of the end of the last major loadDatasets).
<br>This change is for Roy Mendelssohn. Thanks to Bob Simons for this code change.
<br>
<li>IMPROVED: GenerateDatasetsXml now changes to CF-1.10 (was CF-1.6) in the "Conventions" attributes.
<br>Thanks to Bob Simons for this code change.
<br>
<li>Some small changes and bug fixes.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.22" href="#changes2.22" rel="bookmark">Changes</a>
in ERDDAP™ version 2.22 (released 2022-12-08)</h2>
Note that this release was done by Bob Simons, thereby showing that he
is still around and active during the transition to his successor.
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>(None)
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: nothing.
<br>
<li>SECURITY BUG FIX: There was a Cross Site Scripting-related bug in
the code for the language selection drop down. Thanks to NOAA security
scans for catching this. This shows that NOAA security is actively
and routinely looking for security weaknesses in ERDDAP.
<br>
<li>SECURITY FIX: The many libraries used by ERDDAP™ were updated,
as usual, as part of this release.
This time, this included updating the PostgreSQL driver (which had a security bug) to 42.5.1.
<br>
<li>IMPROVED: More small changes to ERDDAP's memory management system
should reduce the chance of a given request failing due to lack of
available memory.
<br>
<li>Some small changes and bug fixes.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.21" href="#changes2.21" rel="bookmark">Changes</a>
in ERDDAP™ version 2.21 (released 2022-10-09)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>(None)
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: For Java 17, you shouldn't use <kbd>-d64</kbd> in JAVA_OPTS in setenv.bat or setenv.sh.
So if it is there, please remove it.
I think that 64 bit mode is now selected when you download a 64 bit version of Java.
Thanks to Sam Woodman.
<br>
<li>BUG FIX: Sometimes, the new email system attempted to log in too often,
which caused Google Email servers to reject all future log in attempts. Now,
the email system avoids this and related problems.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.20" href="#changes2.20" rel="bookmark">Changes</a>
in ERDDAP™ version 2.20 (released 2022-09-30)</h2>
<ul>
<li><b>Don't use v2.20. It is flawed.</b> But administrators still need to do the TO DO
items listed below when upgrading to v2.21+.
<br>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>(None)
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>IMPROVED: We re-enabled the old memory management system (Math2.ensureMemoryAvailable) and modified
the new memory management system (EDStatic.shedThisRequest) to work better with it.
See <a rel="help"
href="https://erddap.github.io/setup.html#MemoryStatus"
>Memory Status</a> for details.
<br>
<li>CHANGED: The default for <ipAddressMaxRequests> in datasets.xml
was increased from 7 to 15. It's clear that some legitimate WMS clients
can generate more than 7 simultaneous requests.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.19" href="#changes2.19" rel="bookmark">Changes</a>
in ERDDAP™ version 2.19 (released 2022-09-01)</h2>
<ul>
<li><b>Don't use v2.19. It is flawed.</b> But administrators still need to do the TO DO
items listed below when upgrading to v2.20+.
<br>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>NEW: There is a new server-side function, orderByDescending, which works like
orderBy, but sorts in descending order.
Thanks to Adam Leadbetter.
<br>
<li>IMPROVED: Now, graphs (but not maps) will expand to fill the available space on the canvas,
i.e., space not used by the legend.
You can get tall graphs, square graphs, or wide graphs by adding and manipulating the
<kbd>&.size=<i>width</i>|<i>height</i></kbd> parameter (where width and height specify the
size of the canvas, in pixels) on the request URL. (This is not an option on the
.graph web page. You have to add it to the URL manually.)
If you don't specify the &.size parameter, requests for .smallPng, .png, .largePng,
.smallPdf, .pdf, and .large.pdf have predefined canvas sizes, so your graph
will expand to fill the available space, but will usually be roughly square.
Thanks to Bob Fleming.
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: ERDDAP™ now requires Java 17 and the related Tomcat 10.
You must follow the ERDDAP™ installation instructions (or the equivalent
e.g., for Docker) to install Java 17 and Tomcat 10 and copy your [tomcat]/content directory
from your Tomcat 8 installation into the new [tomcat] directory.
There are no other changes that you need to make to your ERDDAP
installation related to this change. In other words, ERDDAP™ works
as it did before.
<p>Don't forget to make the ERDDAP-related changes to Tomcat's server.xml and context.xml
when you upgrade Tomcat. See ERDDAP's
<a rel="help" href="https://erddap.github.io/setup.html#tomcat"
>Tomcat installation instructions</a>.
<p>My impression of Java 17 is that it prefers more processing power and memory
for long-running, larger applications like ERDDAP™, so it works slightly slower
than Java 8 with low power computers (e.g., 2 cores and minimal RAM)
and works slightly faster than Java 8 with higher power computers
(e.g., 4+ cores and plentiful RAM).
So if you see poor performance, use programs like Linux's
<a rel="bookmark" href="https://www.howtogeek.com/668986/how-to-use-the-linux-top-command-and-understand-its-output/">top<img
src="../images/external.png" alt=" (external link)"
title="This link to an external website does not constitute an endorsement."></a>
to check resource usage and consider giving ERDDAP™ more resources, notably more memory.
Memory is cheap! Most phones have more processors and memory than
the servers that some of you are using to run ERDDAP!
<br>Thanks to Erin Turnbull.
<br>
<li>TO DO: If you use ERDDAP™ to access Cassandra, for Cassandra, you need to keep using the version
of Java that you were using for running the Cassandra.
Just switch to Java 17 for running Tomcat+ERDDAP.
<br>
<li>TO DO: Recommended: If your server's CPU has 4+ cores and 8+ GB of RAM,
consider changing to these settings in your datasets.xml file:<kbd>
<br> <nGridThreads>3</nGridThreads>
<br> <nTableThreads>3</nTableThreads> </kbd>
<br>If your server has fewer resources, stick to "1" for both of those settings.
<br>The nThreads systems for EDDGridFromFiles and EDDTableFromFiles were significantly improved.
These changes led to a huge speed improvement (e.g., 2X speedup when nThreads is set to 2 or more)
for the most challenging requests (when a large number of files must be processed to gather the results).
Some related changes from Chris John will also lead to a general speedup throughout ERDDAP.
The code for these changes was contributed by Chris John. Thank you, Chris!
<br>
<li>WARNING: hyphens in datasetID's are deprecated and no longer supported (although technically still allowed).
They will probably be disallowed in the next release.
If you use hyphens, switch to underscores now to avoid trouble.
If you make the change now, it's at your own speed. If you wait till the next release,
you'll be in a panic and have to deal with it that day.
<br>
<li>NEW: Now, for .htmlTable data responses,
if the data in a String cell contains <kbd>data:image/png;base64,</kbd> followed by a base64 encoded .png image,
ERDDAP™ will display an icon (so the user can see the image if they hover over it)
and buttons to save the text or the image to the clipboard.
Thanks to Marco Alba (who contributed the code) and Bob Simons (who modified it slightly).
<br>
<li>NEW: -doNotAddStandardNames
<br>If you include <kbd>-doNotAddStandardNames</kbd>
as a command line parameter when you run generateDatasetsXml,
generateDatasetsXml will not add <kbd>standard_name</kbd> to the <kbd>addAttributes</kbd>
for any variables other than variables named <kbd>latitude, longitude, altitude, depth</kbd> or
<kbd>time</kbd> (which have obvious standard_names).
This can be useful if you are using the output from generateDatasetsXml directly in
ERDDAP™ without editing the output, because generateDatasetsXml often guesses
standard_names incorrectly. (Note that we always recommend that you
do edit the output before using it in ERDDAP.) Using this parameter
will have other minor related effects because the guessed standard_name
is often used for other purposes, e.g., to create a new long_name,
and to create the colorBar settings.
Thanks to Kevin O'Brien.
<br>
<li>NEW: You can now put <kbd><updateMaxEvents>10</updateMaxEvents></kbd>
in datasets.xml (in with the other settings near the top) to change the maximum
number of file changes (default=10) that will be processed by the updateEveryNMillis system.
A larger number (100?) may be useful when it is very important that the dataset be kept always up-to-date.
See the <a rel="help"
href="https://erddap.github.io/setupDatasetsXml.html#updateMaxEvents"
>updateMaxEvents documentation</a>.
Thanks to John Maurer.
<br>
<li>NEW: Added support for global "real_time=true|false" String attribute.
<br>If this is <kbd>false</kbd> (the default) and if the dataset doesn't use <kbd>updateEveryNMillis</kbd>,
ERDDAP™ will cache responses to requests
for file types where the entire file must be created
before ERDDAP™ can begin to send the response to the user and reuse them for
up to about 15 minutes (e.g., .nc, .png).
<br>If this is set to <kbd>true</kbd> or if the dataset does use <kbd>updateEveryNMillis</kbd>,
ERDDAP™ will never cache the response files
and will always return newly created files.
<br>Thanks to John Maurer.
<br>
<li>NEW: Emails are now sent in a separate emailThread. This makes loading datasets
and other actions that generate emails faster because loadDatasets doesn't have to
wait for the email to be sent, which sometimes takes a long time.
The new system can send multiple emails per email session,
thus reducing the number of email server logins and reducing the
risk of those failing because they are too frequent.
There are statistics for the emailThread on the status.html page
and diagnostic messages in log.txt -- look for "emailThread".
Note that a tally of nEmailsPerSession=0, indicates trouble,
i.e., an email session was unable to send any emails.
<br>Thanks to Bob Simons.
<br>
<li>CHANGED: Emails are now sent with slightly different code (because of Java 17
and the change to emailThread). If you have trouble sending emails,
please email erd.data at noaa.gov .
<br>
<li>NEW: Subscription actions that "touch" a remote URL are now handled
in a separate touchThread. This makes loading datasets
and other actions that touch URLs faster because loadDatasets doesn't have to
wait for the touch to be completed, which sometimes takes a long time.
There are statistics for the touchThread on the status.html page
and diagnostic messages in log.txt -- look for "touchThread".
<br>Thanks to Bob Simons.
<br>
<li>NEW: On the status.html page, in the "Major LoadDatasets Time Series",
there is a new "shed" column which indicates the number of requests
which were shed because current ERDDAP™ memory use was too high.
Requests which are shed will return HTTP status code 503 "Service Available".
Those requests weren't necessarily a problem. They just arrived at a busy time.
This was part of a revamp of how ERDDAP™ deals with high memory usage.
<br>
<li>NEW: On Unix/Linux computers, there is now an "OS Info" line on the status.html web page with
current operating system information including CPU load and memory use.
<br>
<li>IMPROVED: Now, when ERDDAP™ is restarted and quickRestart=true, EDDTableFromFiles datasets will reuse
subset.nc and distinct.nc. For some datasets, this greatly decreases the
time to load the dataset (e.g., from 60 seconds to 0.3s).
Along with the new emailThread and taskThread (see above), this should
greatly speed up restarting ERDDAP™ for many ERDDAP™ installations.
Thanks to Ben Adams and John Kerfoot.
<br>
<li>CHANGED: Previously, orphan datasets (datasets that are live in ERDDAP™ but are not in datasets.xml)
were simply noted on status.html and in log.txt after each major loadDatasets. Now, they are automatically removed
from ERDDAP™ and noted on status.html and in log.txt, and emailed to emailEverythingTo.
So if you want to remove a dataset from ERDDAP™, now all you have to do is remove
its chunk of xml in datasets.xml and it will be removed in the next major loadDatasets.
Thanks to Bob Simons.
<br>
<li>KNOWN BUG in netcdf-java v5.5.2 and v5.5.3: The EDDGridFromThreddsCatalog option in GenerateDatasetsXml
used to work for THREDDS catalogs that include references to datasets in remote
THREDDS catalogs. Now it doesn't. I have reported the problem to the netcdf-java developers.
<br>
<li>BUG FIX: For Docker users setting setup.xml parameters via ERDDAP_<i>paramName</i>:
for int and boolean parameters (e.g., emailSmtpPort), ERDDAP™ was incorrectly looking for just <i>paramName</i>.
Now it looks for <i>ERDDAP_paramName</i>.
Thanks to Alessandro De Donno.
<br>
<li>CHANGE: The ERDDAP™ testing system now uses an automated system to check that
newly created test images are exactly as expected. Thanks to Chris John for the suggestion
and Bob Simons for the implementation.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.18" href="#changes2.18" rel="bookmark">Changes</a>
in ERDDAP™ version 2.18 (released 2022-02-23)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>NONE
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>BUG FIX: .nc files weren't closed in some circumstances. Now they are.
Thanks to Marco Alba, Roland Schweitzer, John Maurer, and others.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.17" href="#changes2.17" rel="bookmark">Changes</a>
in ERDDAP™ version 2.17 (released 2022-02-16)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>BUG FIX: After changes to the orderBy system a few years ago,
Tabledap's Make A Graph didn't properly handle many queries which used orderBy<i>Xxx</i>.
Now it does.
Thanks to Maurice Libes.
<br>
<li>CHANGE: Previously, ERDDAP™ rejected requests for .transparentPng's when
the latitude and/or longitude values were partly or fully out-of-range.
(ERDDAP™ GitHub Issues #19, posted by Rob Fuller -- thanks for posting that Rob)
Now it returns transparent pixels for any out-of-range areas of the image.
This is useful for many client applications.
The code changes to make this change were done entirely by Chris John.
Thank you very much, Chris!
<br>
<li>CHANGE: Previously, ERDDAP™ rejected griddap requests where the index values
for a given dimension were [high:low].
Now it makes those requests valid by swapping the low and high values.
This solves a longstanding problem for users and for external programs like xtracto
which had to keep track of the few datasets that have latitude values
which range from high to low in order to make request like [(50):(20)]
so that the request in index space was [low:high]. See
https://coastwatch.pfeg.noaa.gov/erddap/griddap/jplAquariusSSS3MonthV5.html
Now, a request like [(20):(50)] for one of these datasets is automatically
interpreted as [(50):(20)].
<br>
<li>CHANGED: .esriAscii requests now trigger a "File : Save As" dialog box in the user's browser.
Thanks to Joel Van Noord.
<br>
<li>BUG FIX: Now, if the longitude variable of a child dataset of a
EDDGridLonPM180 or EDDGridLon0360 dataset
has a valid_min and/or valid_max attribute, they are removed in the
EDDGridLonPM180 or EDDGridLon0360 dataset.
Thanks to Roy Mendelssohn.
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: If you had set <kbd><dataProviderFormActive></kbd> to <kbd>false</kbd>
to temporarily deal with the XSS vulnerability, please set it back to <kbd>true</kbd>.
<br>
<li>SECURITY BUG FIX: Fixed XSS vulnerability in Data Provider Form.
Thanks to Genaro Contreras Gutiérrez.
<br>
<li>BUG FIX: When an AWS S3 dirctory had more than 10000 files, ERDDAP™ threw an "Internal Error".
This is now fixed.
Thanks to Andy Ziegler.
<br>
<li>BUG FIX: EDDGridSideBySide didn't allow to variable's sourceNames in different
child datasets to be the same. Now it does.
Thanks to Joshua Stanford.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.16" href="#changes2.16" rel="bookmark">Changes</a>
in ERDDAP™ version 2.16 (released 2021-12-17)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>CHANGES/BUG FIXES: Numerous small changes to the translation system
thanks to suggestions from language-specific editors.
Thanks to Melanie Abecassis, Marco Alba, Jessy Barrette, Filipe Fernandes,
Etienne Godin, Jennifer Sevadjian, and Mike Smit.
<br>
<li>ADDED a proper disclaimer and attribution for Google Translate,
as required by the terms of Google Translate.
Also, the <html> tag in the HTML for
every web page now properly identifies non-English web pages as
having been machine translated.
Thanks to Mike Smit.
<br>
<li>BUG FIX: The login web pages are now working properly with
different language settings.
Thanks to Mike Smit.
<br>
<li>NEW orderBySum filter. And new Check All and Uncheck All buttons on
EDDGrid Data Access Form web page. Thanks to the code contribution
by Marco Alba.
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: If you have
<br><kbd><questionMarkImageFile>QuestionMark.jpg</questionMarkImageFile></kbd>
<br>in your setup.xml file, you need to either remove the whole tag (recommended, so the default file is used) or change it to:
<br><kbd><questionMarkImageFile>QuestionMark.png</questionMarkImageFile></kbd>
<br>
<li>CHANGE: Just so you know,
<a rel="bookmark" href="https://adoptium.net/?variant=openjdk8">Adoptium<img
src="../images/external.png" alt=" (external link)"
title="This link to an external website does not constitute an endorsement."></a>
has replaced AdoptOpenJDK as the main/recommended source of Java (OpenJDK).
<br>
<li>CHANGE: The log files from ERDDAP™, GenerateDatasetsXml, and DasDds are now
UTF-8, not the computer's default character set.
I did a lot of checking and made a few changes to ensure that ERDDAP™ always
specifies the proper character set when reading or writing all kinds of files, and no
longer (in several cases) relies on the computer's default character set.
This corrected a few mistakes and moved as close as I could to the goal of using UTF-8
for as many file types as possible
(e.g., .log, .xml, .html, .json, .jsonl, .ncHeader).
Note that many older file types are required to use ISO-8859-1
(e.g., OPeNDAP .das, .dds, .csv, .tsv, .nc3, .nccsv, .cpt).
I previously tried to work with the CF group and with Unidata to
add support for UTF-8 in .nc3 files; both were resistant.
<br>
<li>NEW: When downloading files from AWS S3, ERDDAP's cacheFromUrl system
in EDDGridFromFiles and EDDTableFromFiles now uses
the new AWS Transfer Manager to download files via parallelized chunks (thus very fast).
The target throughput is set to 20 Gbps, per file, so this works well with all AWS instance types,
but especially ones which have excellent "Networking Performance".
With this change ERDDAP's cacheFromUrl system now offers comparable speeds
to xarray's approach of parallelized downloads of pre-chunked files,
but without the need to convert the source files from .nc and .hdf into chunked xarray files.
In fact, ERDDAP's system is better if there is a subsequent request to read from the same file,
because ERDDAP™ now has a local copy of the file.
Our community has spent years standardizing on .nc and .hdf files. Now
we don't have to toss that all out just to get good performance when storing
data in AWS S3.
Thanks to Rich Signell.
<br>
<li>CHANGE: searchEngine=Lucene is, for now, deprecated.
It is a complex system which often yields results which are slightly different from
the more desirable behavior of searchEngine=original.
For almost all ERDDAP™ installations, the time savings of Lucene don't offset the differences in results.
Please use searchEngine=original instead if possible.
If that causes problems, please email Bob.
<br>
<li>CHANGE: The Lucene searchEngine now behaves more like the original searchEngine.
There are no longer any cases where lucene thinks a dataset matches and original doesn't.
Also, lucene's rankings now equal original's rankings (because original is now always
used to compute the rankings).
<br>
<li>BUG FIX: Starting in a recent release, ERDDAP™ stopped seeing more than
the first 1000 objects in a given AWS S3 bucket. Now, ERDDAP™ again sees
all of the objects.
Thanks to Andy Ziegler.
<br>
<li>BUG FIX: Now EDDTableAggregateRows removes the actual_range attribute
whenever one or more of the child datasets doesn't ever know its variables' actual_range
(e.g., EDDTableFromDatabase).
Thanks to Erik Geletti.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.15" href="#changes2.15" rel="bookmark">Changes</a>
in ERDDAP™ version 2.15 (released 2021-11-19)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>ERDDAP™ has a new system to let user's specify the language to be used for all web pages.
If an ERDDAP™ installation is set up to use it, the list of languages will appear
in the upper right corner of every web page.
ERDDAP™ URL's from before this version continue to work and always return
English content, as before.
<p>Not all text or all web pages were translated. There were time constraints
on this project that prevented Qi and Bob from getting to 100%.
<p>The obvious question is: why did we put so much effort into this when Chrome
will translate web pages on-the-fly? The answer is: this way, we get much more
control over how the translation is done. Notably, there are lots of words
that shouldn't be translated on the web pages, e.g., the titles and summaries of datasets,
the names of variables,
parameters, units, and organizations. Much of the translation effort was identifying
words and phrases that shouldn't be translated. Also, the machine translations
tended to mangle certain types of HTML markup. Managing the translation
allowed us to minimize this problem.
<p>The translation project was done by Qi Zeng (a Google Summer of Code intern) and Bob Simons
using Google's Translation web service. It was a huge project. Thanks, Qi!
<li>BUG FIX: ERDDAP™ now allows ORCID ID's to have X as last digit.
Thanks to Maurice Libes.
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO:
<ul>
<li>You need to make a few changes related to ERDDAP's new system
to let users specify the language for web pages.
<ul>
<li>On the first line of your setup.xml and datasets.xml files, change to: <kbd>encoding="UTF-8"</kbd>
and change the document's encoding in your text editor so it is saved as a UTF-8 file.
GenerateDatasetsXml now assumes that the datasets.xml is a UTF-8 file.
<li>Programmers who compile ERDDAP:
All of ERDDAP™ .java files should be treated as UTF-8 files by default.
You may need to add "-encoding UTF-8" to the javac command line. (I did.)
<li>To enable this system (strongly recommended), in the <startBodyHtml5>
tag that you specify in datasets.xml, change "&loginInfo;" into "&loginInfo; | &language;"
so that the list of languages appears in the upper right corner of every ERDDAP™ web page.
<li>ERDDAP™ only uses the <startBodyHtml5> tag that you specify in datasets.xml
to specify the HTML content for the banner
at the top of every ERDDAP™ web page, no matter what language the user selects.
If you change that tag to use
<br>"&EasierAccessToScientificData;" instead of "Easier access to scientific data" and
<br>"&BroughtToYouBy;" instead of "Brought to you by",
ERDDAP™ will use translated versions of those phrases in the banner.
<li>Similarly, the new default <theShortDescriptionHtml> in datasets.xml is <pre>
<theShortDescriptionHtml><![CDATA[
<h1>ERDDAP</h1>
&erddapIs;
&thisParticularErddap;
[standardShortDescriptionHtml]
]]></theShortDescriptionHtml></pre>
The last 3 lines of content are things that will be substituted with translated text.
If you convert any of them (notably &thisParticularErddap;) or all of them to explicit text
in datasets.xml (which has priority, if present) or messages.xml,
that text will appear no matter what language the user selects.
This is not perfect, but I figured that few administrators would want to edit
<theShortDescriptionHtml> in 35 different files to provide 35 different translated
versions of that tag.
</ul>
</ul>
<br>
<li>CHANGED: Some errors are now handled slightly differently and so may
be added to the tally of "Failed Requests" on status.html and in the Daily Report Email.
So those numbers may be somewhat larger than before.
<br>
<li>BUG FIX: GenerateDatasetsXml for EDDGridLon0360 and EDDGridLonPM180
now excludes source datasets with datasetID=~".*_LonPM180" and datasetID=~".*_Lon0360", respectively.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.14" href="#changes2.14" rel="bookmark">Changes</a>
in ERDDAP™ version 2.14 (released 2021-07-02)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>(none)
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>NEW: EDDGridLon0360 which makes a gridded dataset with longitude values >=0 and <=360
from a gridded dataset with longitude values >=-180 and <=180.
See the <a rel="help"
href="https://erddap.github.io/setupDatasetsXml.html#EDDGridLon0360"
>EDDGridLon0360 documentation</a>.
Thanks to Dale Robinson.
<br>
<li>NEW: ERDDAP™ administrators can now override any value in setup.xml
via an environment variable named ERDDAP_<i>valueName</i> before running ERDDAP.
For example, use ERDDAP_baseUrl overrides the <baseUrl> value.
This can be handy when deploying ERDDAP™ with a container, as you can
put standard settings in setup.xml and then supply special settings
via environment variables.
If you supply secret information to ERDDAP™ via this method, be sure
to check that the information will remain secret.
ERDDAP™ only reads the environment variables once per startup, in the first second of startup,
so one way to use this is: set the environment variables, start ERDDAP™,
wait until ERDDAP™ is started, then unset the environment variables.
Thanks to Marc Portier.
<br>
<li>IMPROVED: Now, if some files in an EDDTableFrom...Files dataset with a lot of files
have some very long String values,
the dataset will load much faster and respond to requests much faster.
Previously, ERDDAP™ would allocate a lot of space for the min and max String values in the
files which are stored with file information for such datasets.
The resulting file was huge, causing it to be written and read slowly.
Thanks to OBIS.
<br>
<li>IMPROVED: Now, ERDDAP™ does a better job of interpreting unusual and invalid character sequences in CSV files.
Thanks to OBIS.
<br>
<li>FIX: After a year of trouble with Cassandra, I finally successfully installed Cassandra (v2) again
and so was able to rerun the tests with Cassandra v2. So now I can more confidently
state that ERDDAP™ works with Cassandra v2 and v3.
Thanks to ONC.
<br>
</ul>
</ul>
<!-- ****************************************************************************** -->
<h2><a class="selfLink" id="changes2.12" href="#changes2.12" rel="bookmark">Changes</a>
in ERDDAP™ version 2.12 (released 2021-05-14)</h2>
<ul>
<li><strong>New Features and Changes (for users):</strong>
<ul>
<li>BUG FIX: If you're on the subscription blacklist, you now can't request a list of your subscriptions.
<br>
</ul>
<li><strong>Things ERDDAP™ Administrators Need to Know and Do:</strong>
<ul>
<li>TO DO: NEW: system to automatically limit the ability of malicious users
and overly aggressive legitimate users to make a large number of simultaneous requests
which would degrade system performance for other users.
There are 3 new optional tags in datasets.xml which you can/should add right
after <graphBackgroundColor> : <pre>
<ipAddressMaxRequests></ipAddressMaxRequests> <!-- current default=7 -->
<ipAddressMaxRequestsActive></ipAddressMaxRequestsActive> <!-- current default=2 -->
<ipAddressUnlimited></ipAddressUnlimited> <!-- default=empty --> </pre>
For further information, see
<a rel="help"
href="https://erddap.github.io/setupDatasetsXml.html#ipAddressMaxRequests">ipAddressMaxRequests</a>.
ERDDAP™ also now prints the "Number of unique users (since startup)" on the status.html page.
<br>Thanks to the person in China attacking my ERDDAP™ installation.
<br>
<li>CHANGE to Postgresql driver behavior: When I updated the Postgresql driver, the
column names in the table list generated by Postgresql and GenerateDatasetsXml came back all uppercase,
instead of all lowercase, as before. I don't know if that will affect other things
since databases often consider those names to be case insensitive.
My test dataset still works correctly.
But if your dataset stops working with this ERDDAP™ update, this is the possible cause
to pursue first.
<br>
<li>BUG FIX: ERDDAP™ now also handles private AWS S3 files correctly.
There were other related improvements to the handling of AWS S3 files.
Thanks to Michael Gangl and Dylan Pugh.
<br>
<li>NEW: EDDGridFromNcFiles and EDDGridFromNcFilesUnpacked
can now read data from "structures" in .nc4 and .hdf4 files.
To identify a variable that is from a structure, the <sourceName> must use the format:
<kbd><i>fullStructureName</i>|<i>memberName</i></kbd>, for example group1/myStruct|myMember .
Thanks to NRL.
<br>
<li>CHANGED: Now, if current memory usage plus this request is even slightly high,
griddap sets nThreads for this request to 1.
Thus, ERDDAP™ conserves memory when memory is scarce.
Thanks to the person in China attacking my ERDDAP™ installation.
<br>
<li>NEW system to monitor the number of open files
(which includes sockets and some other things, not just files) in Tomcat on Linux computers.
If some files mistakenly never get closed, the number of open files may increase until
it exceeds the maximum allowed and numerous really bad things happen.
So now, on Linux computers (the information isn't available for Windows):
<ul>
<li>There is a new "Open Files" column on the far right of the status.html web page
showing the percent of max files open. On Windows, it just shows "?".
<li>When ERDDAP™ generates that information at the end of each major dataset reload,
it will print to the log.txt file:
<br><kbd> openFileCount=<i>current</i> of max=<i>max</i> %=<i>percent</i></kbd>
<li>If the percentage is >50%, an email is sent to the ERDDAP™ administrator
and the emailEverythingTo email addresses.
</ul>
To find out more, or if you see this problem on your ERDDAP™, see
<a rel="help"
href="https://erddap.github.io/setup.html#TooManyOpenFiles"
>Too Many Open Files</a>.
<br>Thanks to the person in China attacking my ERDDAP™ installation.
<br>
<li>NEW: I added a lot of checking for and handling of "Too many open files", so the
task just stops and the user sees the error message. Data files will no longer be marked as bad
if reading them results in a "Too many open files" error.
<br>
<li>NEW [bigParentDirectory]/badFilesFlag directory:
<br>If you put a file in this directory with a datasetID as the file name (the file contents don't matter),
ERDDAP™ will delete the badFiles.nc file for that dataset (if any) and reload the dataset ASAP.
This causes ERDDAP™ to try again to work with the files previously (erroneously?) marked as bad.
Thanks to Marco Alba.
<br>
<li>CHANGED: At startup, if an EDDGridFrom...Files or EDDTableFrom...Files
dataset initially has 0 files in its list of known valid files (e.g., it's a new dataset),
then ERDDAP™ defers loading it and sets a flag so that it will be loaded
ASAP after the major loadDatasets is finished.
This speeds up the initial startup when there are new datasets.
<br>
<li>CHANGED: FileVisitorDNLS.testAWSS3() and FileVisitorSubdir.testAWSS3();
now use the AWS v2 (not v1) SDK.
So now the Git ERDDAP™ distribution now includes all needed files
and you no longer need to manually add the massive v1 AWS SDK jar file.
<br>
<li>CHANGED: I switched to using Maven to detect/gather dependencies (the .jar files in /lib).
The change to v2 of the AWS SDK necessitated this. It will be needed for
other imported code in the future.
A huge thanks to Kyle Wilcox who provided the pom.xml he created and uses,
which solved several problems for me.
<br>
<li>CHANGED: The classpath parameter (-cp) used in GenerateDatasetXml, DasDds and other small programs
that come with ERDDAP™, and
in the advice to programmers is now much simpler and shouldn't ever change again
since it refers to the directory, not the individual files:
<br><kbd>-cp classes;C:\programs\_tomcat\lib\servlet-api.jar;lib\*</kbd>
<br>(or ':' instead of ';' for Linux and Macs).
<br>(I should have done this years ago when it became an option.)
<br>
<li>NEW: GenerateDatasetsXml has a new utility option: findDuplicateTime
which will search through a collection of gridded .nc (and related) files
to find files with duplicate time values. See
<a rel="help"
href="https://erddap.github.io/setupDatasetsXml.html#findDuplicateTime">findDuplicateTime</a>
<br>
<li>NEW: datasets.xml can now include a <palettes> tag which overrides
the <palettes> tag value from messages.xml (or reverts to the messages.xml
value if it is empty).
This lets you change the list of available palettes while ERDDAP™ is running.
Also, if you have a cptfiles subdirectory in the ERDDAP™ content directory,
ERDDAP™ will copy all the *.cpt files in that directory
into the [tomcat]/webapps/erddap/WEB-INF/cptfiles directory each time ERDDAP™ starts up.
Together, these changes let you add palettes
and have the changes persist when you install a new version of ERDDAP.
See the <a rel="help"
href="https://erddap.github.io/setupDatasetsXml.html#palettes">palettes documentation</a>
<br>Thanks to Jennifer Sevadjian, Melanie Abecassis, and perhaps other CoastWatch people.
<br>
<li>CHANGED: <a rel="help"
href="https://erddap.github.io/setupDatasetsXml.html#slowDownTroubleMillis"><slowDownTroubleMillis></a>
is now used for all failed requests, not just a few types.
<br>
<li>CHANGED: The RunLoadDatasets thread now interrupts the LoadDatasets thread