Metadata store -XML/Harvestor

[ghost]

Essentially I'm trying to figure out how to preserve my metadata store when I shutdown my Geoportal running on AWS.

Related to #9, I'm struggling to figure out where the uploaded metadata is stored (via Catalog and brokers via Harvester). I currently have a AWS EC2/ EFS hosting a docker file that installs Elasticsearch/Tomcat/GeoPortal. I have several metadata files uploaded, and I'm trying to search through the file structure (from root) for anything that resembles "clob", "item", relevant "xml", "index" or, anything resembling the name of the uploaded metadata. Note -only an "index" search comes with anything that even looks worthwhile.

The best I come up with is possible preserving the "nodes" subfolder

Otherwise, I'm lost. Any suggestions? When I stop my EC2 all ephemeral data is lost.

[ghost]

So - I'll just go ahead and answer my own question here. I'm not sure why my first 2 searches for my named metadata didn't come up, but when I uploaded a "SSURGO" metadata file, I was able to find 2 directories with info based on a keyword search:

root@2fbcd74b284e:/# grep -nr 'SSURGO*' .Binary file ./var/lib/elasticsearch/nodes/0/indices/pTtLMnKYRvSYSpCz9m5UEQ/4/translog/translog-3.tlog matchesBinary file ./var/lib/elasticsearch/nodes/0/indices/pTtLMnKYRvSYSpCz9m5UEQ/4/index/_1.cfs matches

When I opened the top file (vim translog/translog-3.tlog), the results looked familiar:

So, it looks like entire "node" subfolder needs to be stored and then referenced in the docker file. I'll do some more checking where the Harvester brokers are stored, and bug the Harvester Github if I can't find an answer.

Thanks!