No description, website, or topics provided.
Clojure XSLT JavaScript HTML Ruby Shell Other
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
access-control-app CMR-4906: Added Cmr-Revision-Id header for ACL update. (#553) Aug 16, 2018
acl-lib CMR-5087 Adds paging to acl-fetcher's search-for-acls call. (#544) Jul 27, 2018
bin CMR-4718: Addressed review feedback. Feb 2, 2018
bootstrap-app CMR-5098 Fixes index rebalancing for old deleted collections. Aug 16, 2018
collection-renderer-lib MMT-1358: Update Preview Gem Jul 20, 2018
common-app-lib CMR-5016: Added CMR-Scroll-Id to allowed and exposed headers. Made ch… Aug 16, 2018
common-lib CMR-2468: Following Chris' feedback, changed to only record number of… Aug 15, 2018
cubby-app CMR-4657: Standardize non-ns test dirs with hyphens. Feb 21, 2018
dev-system CMR-5016: Added CMR-Scroll-Id to allowed and exposed headers. Made ch… Aug 16, 2018
elastic-utils-lib CMR-2078: Added internal mvn/clj repo profile placeholder. Jan 18, 2018
es-spatial-plugin CMR-2078: Added internal mvn/clj repo profile placeholder. Jan 18, 2018
index-set-app CMR-4992: Fix failing test. Jul 5, 2018
indexer-app CMR-5094: Added logging. Aug 17, 2018
ingest-app CMR-2468: Following Chris' feedback, changed to only record number of… Aug 15, 2018
message-queue-lib CMR-4746: Fix services to use transaction ID instead of revision ID w… Feb 16, 2018
metadata-db-app CMR-5094: Added indices on variable_concept_id to variable/service as… Aug 15, 2018
mock-echo-app Cmr 4954 (#535) Jul 18, 2018
oracle-lib CMR-4785: Moved in-memory data store for better parity with Oracle da… Mar 23, 2018
orbits-lib CMR-5007: Fixed the 500 error. (#548) Aug 6, 2018
resources/shell CMR-4718: Addressed review feedback. Feb 2, 2018
search-app Cmr 4954 (#535) Jul 18, 2018
search-relevancy-test CMR-4321: Add more details to the README about the two types of searc… Feb 26, 2018
site-templates CMR-4891: Added missing aliases used in CMR build. Apr 25, 2018
spatial-lib CMR-2078: Added internal mvn/clj repo profile placeholder. Jan 18, 2018
system-int-test Merge pull request #549 from CayvonH/CMR-5016 Aug 17, 2018
transmit-lib CMR-4906: Added Cmr-Revision-Id header for ACL update. (#553) Aug 16, 2018
umm-lib Cmr 4954 (#535) Jul 18, 2018
umm-spec-lib CMR-5039 Adds comments in acl-matcher. Aug 17, 2018
vdd-spatial-viz CMR-2078: Added internal mvn/clj repo profile placeholder. Jan 18, 2018
virtual-product-app CMR-4657: Standardize non-ns test dirs with hyphens. Feb 21, 2018
.gitignore CMR-4718: Cleanup. Feb 2, 2018
.mailmap Add mailmap to fix contributor chart May 16, 2017
CHANGELOG.md Replace instances of REMOVED with something better Apr 11, 2017
CONTRIBUTING.md Flesh out README.md some more Feb 1, 2017
LICENSE CMR-4718: Cleanup. Feb 2, 2018
README.md Typo on string #104 & 143 Jun 6, 2018
project.clj CMR-4718: Cleanup. Feb 2, 2018

README.md

Common Metadata Repository

Visit the CMR at https://earthdata.nasa.gov/about/science-system-description/eosdis-components/common-metadata-repository

About

The Common Metadata Repository (CMR) is an earth science metadata repository for NASA EOSDIS data. The CMR Search API provides access to this metadata.

Client-facing Components

  • Search

  • Ingest

    • Ingest refers to the process of validating, inserting, updating, or deleting metadata in the CMR system and affects only the metadata for the specific Data Partner. The CMR allows Data Partners to ingest metadata records through a RESTful API
    • API Docs: https://cmr.earthdata.nasa.gov/ingest/site/ingest_api_docs.html
  • Access Control

    • Access Control Lists (ACLs) are the mechanism by which users are granted access to perform different operations in the CMR. CMR ACLs follow the same design as ECHO ACLs which in turn are derived from the generic ACL design pattern used in many other systems. At a high level, an ACL is a mapping of actors (subjects) to resources (object) to operations (predicate). For instance, a CMR ACL might specify that all Registered users have READ access to ASTER data or all users in a provider operations group have permissions to ingest data for a particular provider.
    • API Docs: https://cmr.earthdata.nasa.gov/access-control/site/access_control_api_docs.html

Our Development Environment

Prerequisites

  • Java 1.8.0 (a.k.a. JAVA8) only; higher versions are not currently supported.
  • Leiningen (https://leiningen.org) 2.5.1 or above.
    • We've had success with Homebrew and with the install script on the Leiningen website.
  • Ruby (used to support two legacy apps)
  • Maven (https://maven.apache.org/install.html)
    • Mac OS X devs can use brew install maven
    • Linux devs can use sudo apt-get install maven

Obtaining the Code

You can get the CMR source code by cloning the repository from Github:

$ git clone git@github.com:nasa/Common-Metadata-Repository.git cmr

Building and Running the CMR

The CMR is a system consisting of multiple services. The services can be run individually or can run in a single process. Running in a single process makes local development easier because it avoids having to start many different processes. The dev-system project allows the CMR to be run from a single REPL or Jar file. If you're developing a client against the CMR you can build and run the entire CMR with no external dependencies from this Jar file and use that instance for local testing. The sections below contain instructions for running the CMR as a single process or as multiple processes.

Using the cmr CLI Tool

This project has its own tool that is used for everything from initial setup to running builds and tests on the CI/CD infrastructure. In order to use the tool as we do below, be sure to run the following from the top-level CMR directory:

export PATH=$PATH:`pwd`/bin
source resources/shell/cmr-bash-autocomplete

(If you use a system shell not compatible with Bash, we'll accept a PR with auto-complete for it.)

To make this change permanent:

echo "export PATH=\$PATH:`pwd`/bin" >> ~/.profile
echo "source `pwd`/resources/shell/cmr-bash-autocomplete" >> ~/.profile

Oracle Dependencies

Even if you're not going to develop against a local Oracle database, you still need to have the Oracle libraries locally installed to use the CMR.

Here are the steps to do so:

  1. Ensure you have installed on your system the items listed above in the "Prerequisites" section.
  2. Download the Oracle JDBC JAR files into ./oracle-libs/support by following instructions in ./oracle-lib/README.md. (The CMR must have these libraries to build but it does not depend on Oracle DB when running locally. It uses a local in-memory database by default.)
  3. With the JAR files downloaded to the proper location, you're now ready to install them for use by the CMR:" cmr install oracle-libs

Building and Running CMR Dev System in a REPL

  1. cmr setup profile and then update the new ./dev-system/profiles.clj according to personal instructions provided by a CMR code developer.
  2. cmr setup dev
  3. cmr start repl
  4. Once given a Clojure prompt, run (reset)

Note that the reset action could potentially take a while, not only due to the code reloading for a large number of namespaces, but for bootstrapping services as well as starting up worker threads.

Building and Running CMR Dev System from a Jar

Assuming you have already run the above steps (namely cmr setup dev), to build and run the default CMR development system (dev-system) from a .jar file:

  1. cmr build uberjars
  2. cmr start uberjar dev-system

See CMR Development Guide to read about specifying options and setting environment variables

Building and Running separate CMR Applications

The following will build all of the applications but will put each jar into the appropriate target directory for each application. The command shown in step 3 is an example. For the proper command to start up each application, see the Applications section below. Note: Steps 1 and 2 only need to be completed once.

  1. cmr build uberjar APP
  2. cmr run uberjar APP

Where APP is any supported CMR app. You can touble-tap the TAB key on your keyboard to get the cmr tool to show you the list of available apps after entering uberjar in each step above.

Checking Dependencies, Static Analysis, and Tests

There are several lein plugins that have been added to CMR for performing various tasks either at individual subproject levels or at the top-level for all subprojects.

Dependency Versions

To check for up-to-date versions of all project dependencies, you can use cmr test versions PROJ, where PROJ is any CMR sub-project under the top-level directory.

You may run the same command without a project to check for all projects: cmr test versions.

Note that this command fails with the first project that fails. If many subprojects are failing their dependency version checks and you wish to see all of these, you may use your system shell:

for PROJ in `ls -1d */project.clj|xargs dirname`
do
  "Checking $PROJ ..."
  cmr test versions $PROJ
  cd - &> /dev/null
done

Dependency Ambiguities and .jar File Conflicts

To see if the JVM is having problems resolving which version of a dependency to use, you can run cmr test dep-tree PROJ. To perform this against all projects: cmr test dep-trees.

Static Analysis and Linting

To perform static analysis and linting for a project, you can run cmr test lint PROJ. As above with dependency version checking, by not passing a project, you can run for all projects: cmr test lint.

Testing CMR

There are two modes of testing the CMR:

  • From the REPL
  • Utilizing the CI/CD script to run against an Oracle database

For the first, the steps are as follows:

  1. Ensure you have set up your development environment in dev-system
  2. If you have built any .jar files, run cmr clean PROJ (for a given project) or cmr clean to clean all projects.
  3. Start the REPL: cmr start repl
  4. Once in the REPL, start the in-memory services: (reset)
  5. Run the tests: (run-all-tests) or (run-all-tests-future)

Optionally, you could substitute the last step with (run-suites) which uses a third-party tool to display test results more explicitly and with easy copy/paste of tests for running individually (and with easier to read exception messages/stacktraces). Here's an excerpt:

   cmr.system-int-test.ingest.provider-ingest-test
     update-provider-test
       assertion 1 ........................................................ [OK]
       assertion 2 ........................................................ [OK]
       assertion 3 ........................................................ [OK]
       assertion 4 ........................................................ [OK]
     delete-provider-test
       assertion 1 ........................................................ [OK]
       assertion 2 ........................................................ [OK]
       assertion 3 ........................................................ [OK]
       assertion 4 ........................................................ [OK]
       assertion 5 ........................................................ [OK]
       assertion 6 ........................................................ [OK]
       assertion 7 ........................................................ [OK]
       assertion 8 ........................................................ [OK]
       assertion 9 ........................................................ [OK]

For non-terminal based dev, depending upon your IDE/editor, you may have shortcuts available to you for starting/restarting the services and/or running the tests. To find out what these are you can contact a CMR core dev.

In order to run the tests against an Oracle database, it is recommended that you use an Oracle VM built specifically for this purpose. You will also need configuration and authentication information that will be set as environment variables. Be sure to contact a CMR core dev for this information.

Once these are in place, you will be able to run the following:

$ cmr test cicd

If you want to run tests against Oracle, bring up the Oracle VM and execute the following to create the users and run the migrations:

cmr setup db

Then, in the CMR REPL:

user=> (reset :db :external)
...
user=> (run-all-tests)
...

Those tests will take much longer to run than when done with the in-memory database (~25m vs. ~6m). To switch back to using the in-memory database, simply call (reset :db :in-memory).

There is also a different, optional test runner you can use. For more details see the docstring for run-suites in dev-system/dev/user.clj for usage instructions.

Code structure

The CMR is made up of several small services called microservices. These are small purposed-based services that do a small set of things well.

The Microservices

Each microservice has a README file in its root directory, which provides a short overview of the service's functionality. There are a number of main applications, as well as several libraries and support applications.

Applications:

  • access-control-app

    • The mechanism by which users are granted access to perform different operations in the CMR. Also maintains groups and access control rules. Note that users access is provided by either ECHO or URS as an external dependency. The mock-echo application implements both of the necessary interfaces for local testing.
    • Main method: cmr.access_control.runner
  • bootstrap-app

    • Contains APIs for performing various bulk actions in the CMR
    • Main method: cmr.bootstrap.runner
    • See /bootstrap-app/README.md for a list of lein and uberjar commands
  • cubby-app

    • Centralized caching for the CMR. Ideally each application will cache internally, but for situations that require centralized caching, we use Cubby.
    • Main method: cmr.cubby.runner
  • dev-system

    • An app that combines together the separate microservices of the CMR into a single application to make it simpler to develop and run
    • Main method: cmr.dev_system.runner
  • indexer-app

    • This handles indexing collections, granules, and tags in Elasticsearch
    • Main method: cmr.indexer.runner
  • ingest-app

    • The Ingest app is responsible for collaborating with metadata db and indexer components of the CMR system to maintain the lifecycle of concepts coming into the system
    • Main method: cmr.ingest.runner
  • search-app

    • Provides a public search API for concepts in the CMR
    • Main method: cmr.search.runner
  • search-relevancy-test

    • Tests to measure and report the effectiveness of CMR's search relevancy algorithm
  • system-int-test

    • Black-box, system-level tests to ensure functionality of the CMR
  • virtual-product-app

    • Adds the concept of Virtual Products to the CMR. In short: Virtual Products represent products at a data provider that are generated on demand from users when they are ordered or downloaded through a URL
    • Main method: cmr.virtual_product.runner
  • index-set-app

    • An application that maintains the set of indexes in elasticsearch for multiple concept types
    • Main method: cmr.index_set.runner
  • metadata-db-app

    • A database that maintains revisioned copies of metadata for the CMR
    • Main method: cmr.metadata_db.runner
  • mock-echo-app

    • This mocks out the ECHO REST API and the URS API as well. Its purpose is to make it easier to integration test the CMR system without having to run a full instance of ECHO. It won't mock it perfectly or completely. It will only implement the minimum necessary to enable integration testing.
    • Main method: cmr.mock_echo.runner

Libraries:

  • acl-lib

    • Contains utilities for retrieving and working with ACLs
  • common-app-lib

    • Contains utilities used within multiple CMR applications
  • common-lib

    • Provides common utility code for CMR projects
  • elastic-utils-lib

    • Most interfacing with Elasticsearch is done through this application
  • es-spatial-plugin

    • An Elasticsearch plugin that enables spatial search entirely within elastic
  • oracle-lib

    • Contains utilities for connecting to and manipulating data in Oracle
  • orbits-lib

    • Clojure wrapper of a Ruby implementation of the Backtrack Orbit Search Algorithm (BOSA)
  • message-queue-lib

    • A library for interfacing with RabbitMQ, AWS SQS, and an in-memory message queue
  • spatial-lib

    • The spatial libraries provide utilities for working with spatial areas in the CMR
  • transmit-lib

    • The Transmit Library defines functions for invoking CMR services
  • umm-lib

    • This is the old source of UMM schemas and translation code. Since the advent of umm-spec-lib it is actively in the process of being removed
  • umm-spec-lib

    • The UMM Spec lib contains JSON schemas that defined the Unified Metadata Model, as well as mappings to other supported formats, and code to migrate collections between any supported formats.

Further Reading

License

Copyright © 2014-2018 United States Government as represented by the Administrator of the National Aeronautics and Space Administration. All Rights Reserved.