Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362012-05-17T21:09:46ZEcoinformatics Redmine
Redmine Bug #5610 (Resolved): check that build system can handle a "0.9.0" releasehttps://projects.ecoinformatics.org/ecoinfo/issues/56102012-05-17T21:09:46ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>We discussed releasing sensor-view as sensor-view-0.9.0 initially to signify beta status. Check if this is currently possible.</p> Bug #5606 (Resolved): move plotting classes to sensor-viewhttps://projects.ecoinformatics.org/ecoinfo/issues/56062012-05-10T22:07:34ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>THere are plotting classes in runtimemonitor and sensor-view, it seems to me they should all be in one place, sensor-view.</p> Bug #5605 (Resolved): Kepler scheduler doen't work for local workflows in windows oshttps://projects.ecoinformatics.org/ecoinfo/issues/56052012-05-09T22:30:44ZJing Taotao@nceas.ucsb.edu
<p>Before the sensor-view workshop, derik found that the Kepler scheduler doen't work for local workflows in windows. It works in mac and linux.</p> Bug #5602 (Resolved): Timezone confusion using the schedulerhttps://projects.ecoinformatics.org/ecoinfo/issues/56022012-05-07T23:52:44ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>It appears at least some parts of the Workflow Scheduling system don't account for different timezones. When you schedule a workflow in Kepler, you don't specify timezone, so presumably this is local time. I think the server also uses local time, and so this requires you to know where the server resides, and schedule using its time. Instead we should probably translate to GMT, and expose timezone codes in the gui.</p> Bug #5580 (Resolved): create sensor-view module documentationhttps://projects.ecoinformatics.org/ecoinfo/issues/55802012-03-26T20:52:03ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Write sensor-view module documentation.</p> Bug #5417 (Resolved): changes to a sensor's serial-number param do not get written to config filehttps://projects.ecoinformatics.org/ecoinfo/issues/54172011-05-31T20:20:54ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>If you Configure Actor a sensor actor and change it's serial-number, this change is lost on SPAN restart because it is not written to the config file.</p> Bug #5408 (Resolved): dataturbine actor should reuse connectionhttps://projects.ecoinformatics.org/ecoinfo/issues/54082011-05-21T00:43:51ZDaniel Crawldanielcrawl@gmail.com
<p>When running the archival workflow, the dt actor reconnects to the dt server many times. Since the same dt server is used, the actor should reuse the same connection until the workflow finishes.</p> Bug #5404 (Resolved): change-sampling.kar demo no longer workshttps://projects.ecoinformatics.org/ecoinfo/issues/54042011-05-13T19:06:10ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>This demo seems to get stuck at preinitializing.</p> Bug #5403 (Resolved): Improve the performance of the workflow archiving data from dataturbine to ...https://projects.ecoinformatics.org/ecoinfo/issues/54032011-05-12T21:29:26ZJing Taotao@nceas.ucsb.edu
<p>Now the workflow iterates each sensor sequentially to archiving data. Probably we can parallel the process.</p> Bug #5390 (Resolved): EML created by DataTurbineToMetacat workflow should have the spatial coveragehttps://projects.ecoinformatics.org/ecoinfo/issues/53902011-05-03T22:29:33ZJing Taotao@nceas.ucsb.edu
<p>Current EML document generated by the worklfow only have the temporal coverage. We need to add spatial coverage as well.</p> Bug #5379 (Resolved): SPAN - add writing metadata changes to pending fileshttps://projects.ecoinformatics.org/ecoinfo/issues/53792011-04-13T22:58:56ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>If spanToDT crashes, SPAN will keep running, writing data to files in the pending folder, which are successfully picked up on spanToDT restart. It would be great if SPAN also wrote metadata changes. We'd be able to completely recover from a spanToDT crash then.</p> Bug #5378 (Resolved): add rety logic to SpanTodt, attempt to recover if RBNB goes downhttps://projects.ecoinformatics.org/ecoinfo/issues/53782011-04-13T22:45:16ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>We've found that requesting a large amount of data from RBNB on a gumstix, at least via rdv (1.9.0 and 2.2.1), can get the server into a messed up state, requiring a restart, and the RBNB files written to disk will become corrupt such that a restart of rbnb fails to load the old data (yikes - we'll work on a replication procedure and submit these bug(s) to the rbnb list). When this happen SpanTodt locks up. Even when rbnb is successfully restarted (e.g. by deleting rbnb files on disk and starting fresh), SpanTodt stays locked. We should add some retry logic so spanToDT can attempt to persevere through tough times.</p> Bug #5375 (Resolved): data not sent out after accumulating a large number of files in pending/https://projects.ecoinformatics.org/ecoinfo/issues/53752011-04-05T19:55:39ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>There seem to be a few problems associated with letting a large number of files build up in spandata/pending/ (files buffer here until the next connection to 55056, at which point they should all be sent out).</p>
<p>1) The main issue is that after N files (hundreds?) have built up in pending, connecting to 55056 does not cause the send to occur. At least this has happened twice to me at span r24. Connecting to 55056 w/ spanToDT or telnet makes no difference.</p>
<p>2) Errors in the log due to references to filenames that are simply the file extension:<br />------<br /> 50216 <<< Dequeue Data:[2011-04-05T11:40:51.670Z CR800_sq311_2 0.0000000000E+00<br /> 50217 ]<br /> 50218 current file will expired in 0 second(s)<br /> 50219 waiting pending dir become available...<br /> 50220 save data from [./spandata/current.csv] to [./spandata/pending/4D9AFFC0.csv]<br /> 50221 file: ./spandata/pending/.csv will be sent<br /> 50222 cannot open file: ./spandata/pending/.csv<br /> 50223 current file will expired in 25 second(s)<br /> 50224 current file will expired in 20 second(s)<br /> 50225 file: ./spandata/pending/.csv will be sent<br /> 50226 cannot open file: ./spandata/pending/.csv<br /> 50227 current file will expired in 15 second(s)<br /> 50228 current file will expired in 10 second(s)<br /> 50229 file: ./spandata/pending/.csv will be sent<br /> 50230 cannot open file: ./spandata/pending/.csv<br /> 50231 current file will expired in 5 second(s)<br /> 50232 >>> Enqueue Data: 2011-04-05T11:41:21.670Z CR800_Batt_Volt 1.3305263519E+01<br />----<br />Interestingly, pending/.csv never gets written afaict. Also, despite this, it doesn't look like data is lost, e.g. 4D9AFFC0.csv does exist and contains data.</p>
<p>3) At a certain point expiration times go increasingly negative. E.g.:<br />cannot open file: ./spandata/pending/.csv<br />current file will expired in -29217 second(s)<br />current file will expired in -29222 second(s)<br />file: ./spandata/pending/.csv will be sent</p> Bug #5360 (Resolved): changes to a sensor parameter might not take effect until next current samplehttps://projects.ecoinformatics.org/ecoinfo/issues/53602011-03-30T20:27:10ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>if you change e.g. the sampling rate of a sensor, the sensor won't immediately change to the new rate, it will change the next time it samples. This can mean having to wait a long time to see the change. Also, we're not sure if you actually have to wait for the next sample to come through for the change to actually take effect.</p> Bug #5359 (Resolved): SpanToDT NoClassDefFoundErrorhttps://projects.ecoinformatics.org/ecoinfo/issues/53592011-03-29T23:44:54ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>My guess is this is related to a limitation on number of control connections to SPAN. <br />I'll try to narrow it down, but for now:</p>
<p><strong>) Started running SpanToDT on gumstix. tail -f log, looks fine.<br /></strong>) Import sensor site with Kepler, try to execute.<br />*) SpanTODT gives this error, and seems to stop writing data (though flush messages keep coming out)</p>
<p>Exception in thread "Thread-1" java.lang.NoClassDefFoundError: ptolemy/util/MessageHandler<br /> at util.SpanToDT$WriteToDataTurbineThread.run(SpanToDT.java:384)<br />Caused by: java.lang.ClassNotFoundException: ptolemy.util.MessageHandler not found in java.lang.ClassLoader$1{urls=[file:/home/ubuntu/spanTodt/../dataturbine/bin/rbnb-3.2b5.jar,file:/home/ubuntu/spanTodt/sensorview-utils-1.0.jar], parent=null}<br /> at java.net.URLClassLoader.findClass(URLClassLoader.java:529)<br /> at java.lang.ClassLoader.loadClass(ClassLoader.java:341)<br /> at java.lang.ClassLoader$1.loadClass(ClassLoader.java:1112)<br /> at java.lang.ClassLoader.loadClass(ClassLoader.java:293)<br /> at util.SpanToDT$WriteToDataTurbineThread.run(SpanToDT.java:384)</p>