Skip to content

Latest commit

 

History

History
296 lines (183 loc) · 15.4 KB

README.md

File metadata and controls

296 lines (183 loc) · 15.4 KB
Raw

DataDrive-CommandLine

DataDrive Command Line client

This project provides command-line applications to interface with Datadrive/OCS. The DataDrive command line supplements tools within the OCS suite by providing a way of listening to OCS events without having to interface with AWS. It is compatible with CSSO or CAM authentication systems (for OCS and AOCS respectively).

Installation

From Github Release

This is the typical way to install the released software.

Please visit https://github.com/NASA-AMMOS/DataDrive-CommandLine/releases to get the latest and past releases. Each zip contains binaries for Linux, Windows and OSX.

From Source

If you are a developer or want to make changes, or try an unreleased development version of the software, you will need to install from the source code.

Prerequisites

  • NodeJS: tested with v14.15.0, versions newer should also work.

Building with Docker (recommended)

  • Install Docker if you do not have it (https://www.docker.com/). This allows you to run apps in a sandbox and avoids a variety of configuration and compatibility issues that can arise from building software directly
  • Run Docker
  • Clone this repository and go to [cloned_folder]
  • Run make build
    • Note: Make sure you have artifactory set as another npm registry by creating or updating ~/.npmrc file with the following content added or appended.
      • @gov.nasa.jpl.m2020.cs3:registry=https://cae-artifactory.jpl.nasa.gov:443/artifactory/api/npm/npm-release-local/
      • @gov.nasa.jpl.ammos.ids:registry=https://artifactory.jpl.nasa.gov/artifactory/api/npm/npm-develop-local/
    • Please reference https://www.jfrog.com/confluence/display/RTF/Npm+Registry for a more detailed read up.
  • A ZIP of built binaries will be located in [cloned_folder]/dist. Unzip that file and you will get folders for each supported platform.

Building with NPM

  • Clone this repository and go to [cloned_folder].
  • Run make package
    • Note: Make sure you have artifactory set as another npm registry by creating or updating ~/.npmrc file with the following content added or appended.
      • @gov.nasa.jpl.m2020.cs3:registry=https://cae-artifactory.jpl.nasa.gov:443/artifactory/api/npm/npm-release-local/
      • @gov.nasa.jpl.ammos.ids:registry=https://artifactory.jpl.nasa.gov/artifactory/api/npm/npm-develop-local/
    • Please reference https://www.jfrog.com/confluence/display/RTF/Npm+Registry for a more detailed read up.
  • Built binary will be located in [cloned_folder]/deployment/dist

Getting Started

Getting an authentication token (Logging in)

  • For an OCS (CSSO) environment (such as M2020), download credss from the mission-support area, or one of the latest releases and follow the instructions
  • For an AOCS (CAM) environment, cd into the etc directory of the cloned repository, and run request-ssotoken.sh
    • Example: ./request-ssotoken.sh -login_method=LDAP_CLI -access_manager_uri=https://cam.jpl.nasa.gov/cam where access_manager_uri may be different
    • Run request-ssotoken.sh with no arguments to see the help..

Configure CLI

To configure the CLI, please run the following command:

  • ddrv config --help will show the help screen and available options

  • The basic required configuration is the Datadrive Middleware server and the PEP server, ex: ./ddrv config -d [datadrive_middleware_hostname] -p [pep_hostname]

  • Other typical options include a custom log path and the time interval to roll the log files: ./ddrv config --dd-host datadrive-middle-dev.dev.m20.jpl.nasa.gov --pep-host data.dev.m20.jpl.nasa.gov --logdir log_output_folder --log-date-pattern daily

    The command above will create a configuration JSON file in ~/.datadrive/datadrive.json. Note: You should only need to run this once unless ~/.datadrive/datadrive.json file is deleted or you are using the ddrv CLI with multiple environments.

  • Your mission-support team will typically give you the server names, or potentially even a configuration file

  • Check your configuration by running ddrv config with no arguments

Basic Subscriptions

To listen for new files that have been added in a specific package and download the file to a specified folder.

  • Ex:
    • ./ddrv subscribe -p [ocs_package_name] -o [output_directory_path_here] -r
      • Note: -r flag is to retain the ocs path
      • Note: --help flag displays help information
    • Note: Underneath the covers, this CLI uses the web socket API from the DataDrive Middleware, which itself polls a SQS queue that receives update events from OCS's SNS.

Subscriptions with Wildcard filtering

Similar to Basic Subscriptions above, but only download files that match a wildcard filter.

  • Ex: ./ddrv subscribe -p [ocs_package_name] -o [output_directory_path_here] -r -f [wildcard_filter]
  • -f flag should be followed by a wildcard filter. For example, nick*.txt will match nick_09-24-2019.txt and not nick_09-24-2019.gif.
  • Basic Subscriptions with no -f flag are essentially running the -f flag with * as its value.

Subscriptions with Regex filtering

Similar to Basic Subscriptions above, but filter for only files that match a regex filter.

  • Ex: ./ddrv subscribe -p [ocs_package_name] -o [output_directory_path_here] -r -x [regex_filter]
  • -x flag should be followed by a regex filter. For example, nick.*\.txt will match nick_09-24-2019.txt and not nick_09-24-2019.gif.
  • The regex filter uses ElasticSearch's regex syntax and does not take anchor strings indicating beginning ^ or end $. Values put into the filter will always be anchored and the regex filter provided must match the entire string. For more information on the regex expression syntax, please visit https://www.elastic.co/guide/en/elasticsearch/reference/6.4/query-dsl-regexp-query.html#regexp-syntax.
  • The regex filter requires a "/" when matching a path, otherwise it will match the expression against the file basename

Skip Unchanged

-S or --skip-unchanged: Only download files that are new or have changed.

Uses the OCS index event field s3_object_changed which indicates if associated file is considered changed in S3. If a user specifies the skip-unchanged flag, then for each OCS file event, the CLI will:

  • check if the OCS event indicates that file is unchanged;
  • check if local filesystem already contains a file with same name in the output location.
  • If both S3 unchanged and local file exists, then OCS file event will be skipped.

Note: Skipping only applies to OCS events. When performing Playback, all files are assumed to be changed, even if they actually weren't.

Subscriptions with Saved Search

Instead of defining a ocs_package_name with the -p flag, you can subscribe to a saved search that can be created in the DataDrive UI using the -s flag. The saved search will include a package and some filters.

  • Ex: ./ddrv subscribe -s [saved_search_name] -o [output_directory_path_here] -r
  • Note: The example command uses the -r flag which will retain the ocs path.

Playback events

You can "playback" events that have since happened from a certain date. When "playback" is enabled, the CLI will query OCS for documents (that have not been deleted) since the last saved checkpoint date.

  • -P flag is used to denote the use of the playback option.
  • The checkpoint date is saved in YYYY-MM-DDTHH:mm:ss.SSSZZ format in the <output_dir>/.datadrive/checkpoint.txt file in the specified output folder.

Configuring Logging

By default, all log entries are stored in a single file in the configured log directory (combined.log).

To utilize rolling logfiles, use the --log-date-pattern option with ddrv config. Simple options are monthly, weekly, and daily. You can also specify a custom log rotation value using this format: (https://momentjs.com/docs/#/parsing/string-format/). This format will define the frequency of the log rotation, with the logs rotating on the smallest date/time increment specified in the format.

For example, ddrv config --log-date-pattern YYYYMMDDhhmm will create a new log file every minute, named something like combined-202406131233.log.

By default, logs are archived in a gzip file. To disable this functionality, set the --no_gzip_rolling_logs option when running ddrv config.

The --log-date-pattern, --no_gzip_rolling_logs, and --logdir options can also be used with the ddrv subscribe command to determine the logging options for that particular subscription session.

Subscriptions that auto-execute a command on each downloaded file

You can run a single shell command that will be called for every file downloaded. You can use this to automatically run post-processing, for example.

  • Ex: ./ddrv subscribe -p [ocs_package_name] -o [output_directory_path_here] -r -x [regex_filter] --callback 'command_to_execute'
  • --callback flag's value should be in single quotes
  • --callback flag's value is an shell command and can use these special variables: $FILENAME $SRC_PATH $DEST_PATH
  • for example: ddrv subscribe -if --retain-path -p sample-package -o output_folder -f 'filter-string*' -P --callback 'echo $FILENAME $SRC_PATH $DEST_PATH'

Unit Test

To run unit tests, go to src/ folder and run the command npm test. Note: To avoid confusion, the commands below is for manual testing without packaging files into a single binary.

Manual Tests

Please note, these test cases are for developers performing manual test from the source code. For others that are performing these tests via the published binary, you would start commands below with ddrv instead of node ddrv.js.

Test 1 - Subscriptions Only

Steps

  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp -p DataDrive-Dev-Congruent-1 or ./ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1
  • Upload a file into the specified package in the example above "DataDrive-Dev-Congruent-1"

Expected Results

  • File should be downloaded into the specified folder. In example above, the tmp folder.
  • A date should be written into the <output_dir>/.datadrive/checkpoint.txt file under the specified folder.
    • Ex: 2019-09-08T11:51:28.000-0700

Test 2 - Subscriptions and Playback Events

Steps

  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o /tmp -p DataDrive-Dev-Congruent-1 -P or ./ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1 -P

Expected Results

  • Files since the date listed in <output_dir>/.datadrive/checkpoint.txt will be downloaded into the specified folder.
  • The CLI will now be listening for new files uploaded into the specified package.
  • Reference Subscription Only for expected results for file subscriptions.

Test 3 - Playback Events on Deleted File

Steps

  • Upload files into specified package then delete these files.
  • Make sure <output_dir>/.datadrive/checkpoint.txt file has a date/time that is earlier than the date/time for uploaded files above.
  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp -p DataDrive-Dev-Congruent-1 -P or ./ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1 -P

Expected Results

  • Files uploaded above will NOT be downloaded by the CLI.

Test 4 - Adjusting checkpoint.txt

Steps

  • Upload files into specified package.
  • Go into <output_dir>/.datadrive/checkpoint.txt file and adjust the date to an earlier time before uploaded files.
    • Note: The time format must be in a valid format such as... Ex: 2019-09-08T11:51:28.000-0700.
  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp -p DataDrive-Dev-Congruent-1 -P or ./ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1 -P

Expected Results

  • Files since given date will be downloaded.

Test 5 - Same file name

Steps

  • CLI tries to download a file with the same file name from the middleware (even if the contents are different)

Expected Results

  • An error will be given as the file already exists.

Test 6 - Wildcard Filter

Steps

  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp -p DataDrive-Dev-Congruent-1 -f 'msl*.txt' or ./ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1 -f 'msl*.txt'
  • Upload a file with a file name "msl_sol_123.txt".
  • Upload a file with a file name "europa_day_2345.txt".

Expected Results

  • "msl_sol_123.txt" file will be downloaded to the tmp directory.

Cleanup

  • Delete "msl_sol_123.txt" from both DataDrive web app and files in the tmp directory.

Test 7 - Regex Filter

Steps

  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp -p DataDrive-Dev-Congruent-1 -x 'msl.*\.txt' or ./ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1 -x 'msl.*\.txt'
  • Upload a file with a file name "msl_sol_123.txt".
  • Upload a file with a file name "europa_day_2345.txt".

Expected Results

  • "msl_sol_123.txt" file will be downloaded to the tmp directory.

Cleanup

  • Delete "msl_sol_123.txt" from both DataDrive web app and files in the tmp directory.

Test 8 - Regex Filter and Wildcard Filter

Steps

  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp -p DataDrive-Dev-Congruent-1 -x 'msl.*\.txt' -f 'msl*.txt' or ddrv subscribe -o tmp -p DataDrive-Dev-Congruent-1 -x 'msl.*\.txt' -f 'msl*.txt'

Expected Results

  • An error will be returned to the user as you cannot specify both regex and wildcard filter together.

Test9 - Test out subscripts for a saved search

Steps

  • Ensure that there is a folder called tmp_ss created where you will be downloading file.
  • Ensure that a saved search has been created. This can be done using the DataDrive UI. Create a saved search that will filter based on a directory path in ocs_path field and with value of /jeff
  • Run the command line such as the example below. Note: the -s flag is for the saved search name created above.
    • Ex: node ddrv.js subscribe -o tmp_ss -s 'jeffliu_saved_search_2' -r
  • Upload a file with a name "seeme123.txt" into /jeff folder.
  • Upload a file with a name "seemenot123.txt" not into /jeff folder.

Expected Results

  • "seeme123.txt" file will be downloaded to the tmp_ss directory under /jeff folder.

Cleanup

  • Delete "seeme123.txt" from files in the tmp_ss directory.
  • Delete "seemenot123.txt" from files in the tmp_ss directory.

Test 10 - Adjusting checkpoint.txt for a saved search

Steps

  • This test case depends on test 9 above. Make sure that you run test 9 first.
  • Go into tmp_ss/.datadrive/checkpoint.txt file and adjust the date to an earlier time before uploaded files in test 9.
    • Note: The time format must be in a valid format such as... Ex: 2019-09-08T11:51:28.000-0700.
  • Run the command line such as the example below.
    • Ex: node ddrv.js subscribe -o tmp_ss -s 'jeffliu_saved_search_2' -P

Expected Results

  • Files since given date will be downloaded. In this case, "seeme123.txt" should be downloaded..

Cleanup

  • Delete "seeme123.txt" from files in the tmp_ss directory.