missing VIIRS-RSBAUTOCAL-OBCIP-HISTORY-AUX

[wzchen]

I try to run "ProSdrViirsRSBAutoCal" algorithm using ADL5 Block2. However, it seems that the " VIIRS-RSBAUTOCAL-OBCIP-HISTORY-AUX" LUT is missing. Could you please point me where I can get it?
Thanks,

[tsimpson]

The VIIRS-RSBAUTOCAL-OBCIP-HISTORY-AUX is supposed to be an optional input for ProSdrViirsRSBAutoCal. Look in the config guide, $ADL_HOME/cfg/<spacecraft>/ProSdrViirsRSBAutoCal_CFG.xml, where <spacecraft> is NPP or J01 depending on what data you are trying to process. In the Inputs group, there should be a group named ObcIpHistoryAux with two config items: OfficialShortName_1 with value VIIRS-RSBAUTOCAL-OBCIP-HISTORY-AUX, and Mode with value Optional.

If the Mode config item is not there, add these lines to the ObcIpHistoryAux group and try the algorithm again:

<config>
<name>Mode</name>
<configValue>Optional</configValue>
</config>

If the algorithm still fails, it might help if I could take a look at the log file it produces.

Hope this helps ... please let me know if you have other questions or comments.

[wzchen]

Hi Todd,

I just checked my config file. It is Optional. Please check the log file in the attachment.
Thanks,

[wzchen]

The log said that the file VIIRS-RSBAUTOCAL-HISTORY has been successfully wrote to file. Why can I not see it? However, I can find VIIRS-RSBAUTOCAL-OBCIP-HISTORY file.

2015-07-16 19:30:08.002000Z DBG_HIGH DmCoreCacheInterface.cpp(802)[ProSdrViirsRSBA (47979491081760)]: tid-47979491081760 unmap() successfully wrote file: /data/data096/weizhong/run_viirs_j1_geo/data/output/withMetadata/ProSdrViirsRSBAutoCalOutputs/20120121-J1_VSDR_base_rsb/d20120121_t1129372_e1131025_b00000/55a8063e-c3471-020ab506-32bb3fa5.VIIRS-RSBAUTOCAL-HISTORY-AUX_j01_20150716193006Z_20120121000000Z_ee00000000000000Z_-_null_dev_all-_dev.bin address_==0xc769d20 length_==2266960

Thanks,

[kbisanz]

Looking at the log file around lines 33-38, there are messages that sound as if they could indicate an issue with processing the data.

The log file does not contain much detail because the log level is DBG_HIGH (for only high level details). Try setting the log level to DBG_LOW (for low level detail) and rerunning. You will get many more lines of log messages. Hopefully we can find something useful in them. You could also try DBG_MED for something in between. If the log file is large you may have to gzip it before uploading it.

I am not sure why it said it successfully wrote the file but the file is not present. I am guessing that because the algorithm failed, the output was removed. If the next log file has more detail we might see the remove happening.

[wzchen]

The ADL can be built no problem if I just use the script "buildAdl.ksh" without any arguments. However, as long as I changed debug level to 2 or 3 by using the following command:
./buildAdl.ksh -DiagnosticMode 3
The ADL failed compiling. Here is the log file.
Thanks,

[kbisanz]

I believe you're confusing debug logging with diagnostic code. Diagnostic code (which you don't want in this case) is a way to put code into the algorithm which you only want to execute if diagnostic mode has been enabled. For example, you could perform extra processing and output values from intermediate calculations. Diagnostic mode is intertwined deeply in the code and requires a full rebuild with diagnostic code enabled (which you did perform with -DiagnosticMode 3). That build should have worked but failed for some reason. I don't see the error in the log file. I am guessing you redirect only stdout (and not stderr) into that file. The build error is likely output to stderr. However, we don't need diagnostic mode to investigate your issue. So you need to rebuild ADL again like normal (without the -DiagnosticMode 3).

Debug logging outputs only messages to a log file. It cannot perform extra processing like diagnostic mode. Debug level is controlled by the <debugLevel> tag in your LW file. There is no need to rebuild when changing anything in the LW file, including debugLevel. If you're using the command line chain runner there is a flag to change the debug level. If you're using the GUI chain runner there's a combo to select the debug level. After the recompile (to fix the issue with a diagnostic compile), change the debug level to be DBG_LOW in the LW file or as appropriate with one of the chain runners. There will then be much more info in the algorithm log file.

In summary:
--Rebuild ADL like normal (to fix issues when you compiled with diagnostic mode)
--Enable DBG_LOW
--Run the algorithm

[wzchen]

Here is the new log file with DEBUG LOW.
Thanks,

[kbisanz]

The original DBG_HIGH log file differs from the newly posted DBG_LOW log file, even in regards to the DBG_HIGH messages. In other words the algorithm was doing different things between the two log files. I believe that RSB AutoCal's outputs can feed back into the next algorithm run. For example the following line is not in the first log file.
2015-07-22 16:47:18.803000Z DBG_HIGH ProCmnMethodAudit.cpp(170)[ProSdrViirsRSBA (47436499185184)]: ProCmnSdrAlgorithm[ProSdrViirsRSBAutoCal]::calculateBoundingBox(J01000078173772) [0x2b24abaf8010] PRO_FAIL - Number of BB latitude and longitude points are invalid. from file ProCmnSdrAlgorithm.cpp, line 374

Can you remove everything from your output directory and try to rerun again with DBG_LOW?

[wzchen]

I deleted the output directory. Then reran it again. All the results from ViirsGeo and ViirsCal were included to "INFTK_DM_ROOT" variable in the different output directories.