|
|---|
This page describes HySDS how it is composed of 6 - including its six main components and this is what they dotheir respective purposes. For a more detailed look at each of these components see the L2 overview.
Jobs:
Tasks with additional robustness, associated working directories, and job handling.
Worker:
Workers are processes that run on the compute instances; workers run the tasks
Verdi
Verdi is composed of workers that pull jobs from the queue and executes them. Once the jobs are completed, the finished products are sent to a storage repository. Additionally, the product metadata is indexed in Elasticsearch.
Note: a “compute fleet” refers to an army of Verdi running in compute workers.
Mozart
Mozart is a resource management service which orchestrates job management across the HySDS hybrid cloud computing architecture.
It manages job queues (RabbitMQ), and job metadata (Elasticsearch). Mozart also hosts information about job status, and the GUI for viewing job status using a faceted view (Figaro).
Factotum
Factotum consists of special workers which handle specific processes that don’t generate data products. This includes non-scaling jobs in the AWS cloud.
GRQ (Geo Region Query)
GRQ catalogs product metadata in Elasticsearch indices. It provides the GeoRegionQuery (GRQ) REST API for publishing and querying the metadata. It also hosts the product faceted view GUI (Tosca) for viewing published products.
Metrics
Displays worker and job metrics. It hosts the Elasticsearch indices for real-time job and worker metrics.
Continuous Integration (CI)
GitHub is checked for updates to containers and publishes them when found.
|
Anyone can join our public Slack channelto learn more about HySDS. JPL employees can join #HySDS-Community
|
JPLers can also ask HySDS questions atStack Overflow Enterprise
|
Have Questions? Ask a HySDS Developer:| Live Search | ||
|---|---|---|
|
