Case Study Server Configuration
This document covers what is necessary to install and run AWIPS EDEX as an archive and case study server (no purging of processed data).
Follow the EDEX Install Instructions including iptables config and an optional SSD mount (for large data volumes).
groupadd fxalpha && useradd -G fxalpha awips mkdir -p /awips2/data_store wget -O /etc/yum.repos.d/awips2.repo https://www.unidata.ucar.edu/software/awips2/doc/el7.repo yum clean all yum groupinstall awips2-server -y
Disable Data Purging
The easiest way to disable data purging is to add an
<exclude>purge.*</exclude> entry in
/awips2/edex/conf/modes/modes.xml so that the purge plugin is not loaded when the EDEX ingest JVM is started:
vi /awips2/edex/conf/modes/modes.xml <?xml version="1.0" encoding="UTF-8" standalone="yes"?> <edexModes> <mode name="ingest"> <exclude>.*request.*</exclude> <exclude>edex-security.xml</exclude> ... <exclude>purge.*</exclude> </mode> ... </edexModes>
Start EDEX without running the LDM, since we do not want current data. Run the following command:
edex start base
Double check everything is running, except the LDM:
edex [edex status] postgres :: running :: pid 43644 pypies :: running :: pid 3557 qpid :: running :: pid 43742 EDEXingest :: running :: pid 6564 44301 44597 EDEXgrib :: running :: pid 6565 44302 44598 EDEXrequest :: running :: pid 6566 44303 44599 ldmadmin :: not running
Ingest Case Study Data
Raw data files of any type can be copied or moved into
/awips2/data_store/ingest/ to be picked up and decoded by EDEX. Most data types are recognized by regular expression matching of the WMO Header or filename.
Individual files can be ingested on the command line with the regex header/pattern supplied as the last argument:
qpidNotify.py /full/path/to/data.file [regex match]
qpidNotify.py /home/awips/uniwisc_U5_132GOES-15_IMG10.7um_4km_20171024_1830.area.png uniwisc qpidNotify.py /awips2/data_store/grid/NAM12/conduit/NAM_CONUS_12km_conduit_20171025_1200Z_F084_TMPK-7.000007.grib2 grib qpidNotify.py /awips2/data_store/radar/FTG_N0Q_20171015_1815 Level3
Viewing Archive Data in CAVE
Because we are installing and configuring a standalone EDEX archive server without real-time LDM data ingest (and with purge disabled), any case study data that is ingested will be the "latest available" to CAVE, and you will see CAVE product menu time fill in with the latest of all data ingested.
However, to display specific time-based data (in case you ingest more than one case study), there are two options:
Set Load Mode to Inventory
In the top-left toolbar change Valid time seq to Inventory.
Now any data product selected from the menus or the Product Browser should prompt you to select the exact time.
Set Data Display Time in CAVE
At the bottom of the CAVE application, double-click the Time: entry to bring up a dialog window where you can set CAVE to a previous time, and choose the option of freezing CAVE at that time or allowing CAVE to "move forward in time" from that position as if it were real-time.