Gerald Manipon edited this page on Mar 21 · 36 revisions

Page Navigation:

Table of Contents

Confidence Level TBD This article has not been reviewed for accuracy, timeliness, or completeness. Check that this information is valid before acting on it.

Initial Cluster Setup

For definitions of terminology used, please refer to our terminology reference.

When installation of the HySDS framework is complete on yourmozartinstance (seeInstallation), we must configure the rest of the cluster instances so that they can talk to each other. We do this using thesdsclicommand on themozartinstance. The idea is that all code and configuration is centralized on themozartinstance and when ready to deploy updates during the development cycle or when upgrading operations, we can push them easily from a single location.

Configure your cluster parameters usingsdscli: Thesdsclirepo was installed on yourmozartinstance duringInstallation. Configure your cluster by running:
Code Block
cd ~ source ~/mozart/bin/activate sds configure
Thesds configurecommand will prompt you for your cluster parameters. A description of the parameters with examples is provided below

*** LOOK OVER THE FIELDS AND HAVE THE VALUES READY BEFORE HAND ***

field	description	example
MOZART_PVT_IP	private IP address of`mozart`instance	100.64.134.201
MOZART_PUB_IP	publicly accessible IP address of`mozart`instance, e.g. Elastic IP; can be the same as MOZART_PVT_IP	64.34.21.123
MOZART_FQDN	publicly resolvable FQDN of`mozart`instance; can be the same as MOZART_PVT_IP	gman-jobs.hysds.net
MOZART_RABBIT_PVT_IP	private IP address of`mozart`rabbitMQ instance (if running rabbitMQ on a different instance); otherwise replicate value from MOZART_PVT_IP	100.64.134.201
MOZART_RABBIT_PUB_IP	publicly accessible IP address of`mozart`rabbitMQ instance (if running rabbitMQ on a different instance); otherwise replicate value from MOZART_PUB_IP	64.34.21.123
MOZART_RABBIT_FQDN	publicly resolvable FQDN of`mozart`rabbitMQ instance (if running rabbitMQ on a different instance); otherwise replicate value from MOZART_FQDN	gman-jobs.hysds.net
MOZART_RABBIT_USER	rabbitMQ user account; default installed by rabbitMQ is`guest`	guest
MOZART_RABBIT_PASSWORD	rabbitMQ user password; default installed by rabbitMQ is`guest`	guest
MOZART_REDIS_PVT_IP	private IP address of`mozart`redis instance (if running redis on a different instance); otherwise replicate value from MOZART_PVT_IP	100.64.134.201
MOZART_REDIS_PUB_IP	publicly accessible IP address of`mozart`redis instance (if running redis on a different instance); otherwise replicate value from MOZART_PUB_IP	64.34.21.123
MOZART_REDIS_FQDN	publicly resolvable FQDN of`mozart`redis instance (if running redis on a different instance); otherwise replicate value from MOZART_FQDN	gman-jobs.hysds.net
MOZART_REDIS_PASSWORD	redis password (if AUTH is configured)	empty string or <redis password>
MOZART_ES_PVT_IP	private IP address of`mozart`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from MOZART_PVT_IP	100.64.134.201
MOZART_ES_PUB_IP	publicly accessible IP address of`mozart`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from MOZART_PUB_IP	64.34.21.123
MOZART_ES_FQDN	publicly resolvable FQDN of`mozart`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from MOZART_FQDN	gman-jobs.hysds.net
OPS_USER	ops account on HySDS cluster instances	ops or hysdsops or swotops
OPS_HOME	ops account home directory on HySDS cluster instances	/home/ops or /data/home/hysdsops
OPS_PASSWORD_HASH	sha224sum password hash for ops user account login to HySDS web interfaces	output of`echo -n mypassword \| sha224sum`
LDAP_GROUPS	comma-separated list of LDAP groups to use for user authentication into HySDS web interfaces	hysds-v2,aria.users,ariamh
KEY_FILENAME	private ssh key to use for logging into other cluster instances; used for deployment via fabric	/home/ops/.ssh/my_cloud_keypair.pem
JENKINS_USER	account on`ci`instance that owns and runs the Jenkins CI server	jenkins
JENKINS_DIR	location of the Jenkins HOME directory (where jobs/ directory is located)	/var/lib/jenkins
METRICS_PVT_IP	private IP address of`metrics`instance	100.64.134.153
METRICS_PUB_IP	publicly accessible IP address of`metrics`instance, e.g. Elastic IP; can be the same as METRICS_PVT_IP	64.34.21.124
METRICS_FQDN	publicly resolvable FQDN of`metrics`instance; can be the same as METRICS_PVT_IP	gman-metrics.hysds.net
METRICS_REDIS_PVT_IP	private IP address of`metrics`redis instance (if running redis on a different instance); otherwise replicate value from METRICS_PVT_IP	100.64.134.153
METRICS_REDIS_PUB_IP	publicly accessible IP address of`metrics`redis instance (if running redis on a different instance); otherwise replicate value from METRICS_PUB_IP	64.34.21.123
METRICS_REDIS_FQDN	publicly resolvable FQDN of`metrics`redis instance (if running redis on a different instance); otherwise replicate value from METRICS_FQDN	gman-metrics.hysds.net
METRICS_REDIS_PASSWORD	redis password (if AUTH is configured)	empty string or <redis password>
METRICS_ES_PVT_IP	private IP address of`metrics`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from METRICS_PVT_IP	100.64.134.153
METRICS_ES_PUB_IP	publicly accessible IP address of`metrics`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from METRICS_PUB_IP	64.34.21.124
METRICS_ES_FQDN	publicly resolvable FQDN of`metrics`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from METRICS_FQDN	gman-metrics.hysds.net
GRQ_PVT_IP	private IP address of`grq`instance	100.64.134.71
GRQ_PUB_IP	publicly accessible IP address of`grq`instance, e.g. Elastic IP; can be the same as GRQ_PVT_IP	64.34.21.125
GRQ_FQDN	publicly resolvable FQDN of`grq`instance; can be the same as GRQ_PVT_IP	gman-grq.hysds.net
GRQ_PORT	port to use for the grq2 REST API	8878
GRQ_ES_PVT_IP	private IP address of`grq`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from GRQ_PVT_IP	100.64.134.71
GRQ_ES_PUB_IP	publicly accessible IP address of`grq`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from GRQ_PUB_IP	64.34.21.125
GRQ_ES_FQDN	publicly resolvable FQDN of`grq`elasticsearch instance (if running elasticsearch on a different instance); otherwise replicate value from GRQ_FQDN	gman-grq.hysds.net
FACTOTUM_PVT_IP	private IP address of`factotum`instance	100.64.134.184
FACTOTUM_PUB_IP	publicly accessible IP address of`factotum`instance, e.g. Elastic IP; can be the same as FACTOTUM_PVT_IP	64.34.21.126
FACTOTUM_FQDN	publicly resolvable FQDN of`factotum`instance; can be the same as FACTOTUM_PVT_IP	gman-factotum.hysds.net
CI_PVT_IP	private IP address of`ci`instance	100.64.134.179
CI_PUB_IP	publicly accessible IP address of`ci`instance, e.g. Elastic IP; can be the same as CI_PVT_IP	64.34.21.127
CI_FQDN	publicly resolvable FQDN of`ci`instance; can be the same as CI_PVT_IP	gman-ci.hysds.net
VERDI_PVT_IP	private IP address of`verdi`instance; if no`verdi`instance, use`ci`instance value for CI_PVT_IP	100.64.134.179
VERDI_PUB_IP	publicly accessible IP address of`verdi`instance, e.g. Elastic IP; if no`verdi`instance, use`ci`instance value for CI_PUB_IP	64.34.21.127
VERDI_FQDN	publicly resolvable FQDN of`verdi`instance; if no`verdi`instance, use`ci`instance value for CI_FQDN	gman-ci.hysds.net
JENKINS_API_USER	Jenkins user account to use for access to Jenkins API	gmanipon
JENKINS_API_KEY	Jenkins user API key to use for access to Jenkins API. Go to an already set up Jenkins web page and click on “People”, your username, then “Configure”. Click on “Show API Token”. Use that token and you username for API_USER.	<api key>
DAV_SERVER	WebDAV server for dataset publication (optional); leave blank if using S3	aria-dav.jpl.nasa.gov
DAV_USER	WebDAV server account with R/W access	ops
DAV_PASSWORD	DAV_USER account password	<password>
DATASET_AWS_ACCESS_KEY	AWS access key for account or role with R/W access to S3 bucket for dataset repository	<access key>
DATASET_AWS_SECRET_KEY	AWS secret key for DATASET_AWS_ACCESS_KEY	<secret key>
DATASET_AWS_REGION	AWS region for S3 bucket for dataset repository	us-west-2
DATASET_S3_ENDPOINT	S3 endpoint for the DATASET_AWS_REGION	s3-us-west-2.amazonaws.com
DATASET_S3_WEBSITE_ENDPOINT	S3 website endpoint for the DATASET_AWS_REGION	s3-website-us-west-2.amazonaws.com
DATASET_BUCKET	bucket name for dataset repository	ops-product-bucket
AWS_ACCESS_KEY	AWS access key for account or role with R/W access to S3 bucket for code/config bundle and docker image repository; can be the same as DATASET_AWS_ACCESS_KEY	<access key>
AWS_SECRET_KEY	AWS secret key for AWS_ACCESS_KEY; can be the same as DATASET_AWS_SECRET_KEY	<secret key>
AWS_REGION	AWS region for S3 bucket for code/config bundle and docker image repository	us-west-2
S3_ENDPOINT	S3 endpoint for the AWS_REGION	s3-us-west-2.amazonaws.com
CODE_BUCKET	bucket name for code/config bundle and docker image repository	ops-code-bucket
VERDI_PRIMER_IMAGE	S3 url to`verdi`docker image in CODE_BUCKET	s3://ops-code-bucket/hysds-verdi-latest.tar.gz
VERDI_TAG	docker tag for`verdi`docker image	latest
VERDI_UID	UID of ops user on`ci`instance; used to sync UID upon docker image creation	1001
VERDI_GID	GID of ops user on`ci`instance; used to sync GID upon docker image creation	1001
QUEUES	space-delimited list of queues to create autoscaling code/config bundles for	"dumby-job_worker-small dumby-job_worker-large"
INSTANCE_TYPES	space-delimited list of instance types to use for the corresponding queue as defined in`QUEUES`	"t2.micro t2.micro"
VENUE	unique tag name to differentiate this HySDS cluster from others	e.g. ops or dev or oasis or test
PROVES_URL	url to PROV-ES server (optional)	https://prov-es.jpl.nasa.gov/beta
PROVES_IMPORT_URL	PROV-ES API url for import of PROV-ES documents (optional)	https://prov-es.jpl.nasa.gov/beta/api/v0.1/prov_es/import/json
DATASETS_CFG	location ofdatasets configurationon workers	/home/ops/verdi/etc/datasets.json
SYSTEM_JOBS_QUEUE	name of queue to use for system jobs	system-jobs-queue
GIT_OAUTH_TOKEN	optional Github OAuth token to use on`ci`instance when checking out code for continuous integration (optional)	<token>

Make sure elasticsearch is up on the mozart and grq instances. You can run the following command to check:

Code Block
curl 'http://<mozart/grq ip>:9200/?pretty'

you should get answer back from ES, something like this:

Code Block

{
 "status" : 200,
 "name" : "Dweller-in-Darkness",
 "cluster_name" : "resource_cluster",
 "version" : {
   "number" : "1.7.3",
   "build_hash" : "05d4530971ef0ea46d0f4fa6ee64dbc8df659682",
   "build_timestamp" : "2015-10-15T09:14:17Z",
   "build_snapshot" : false,
   "lucene_version" : "4.10.4"
 },
 "tagline" : "You Know, for Search"
}

If you can not connect to elastic search, you need to start ElasticSearch in mozart and grq instances:

Code Block
sudo systemctl start elasticsearch

Ensuremozartcomponent can connect to other components over ssh using the configuredKEY_FILENAME. If correctly configured, thesds status allcommand should show that it was able to ssh into each component to check that thesupervisorddaemon was not running like below:

Code Block

sds status
########################################
grq
########################################
----------------------------------------
third-party services
----------------------------------------
[100.64.134.135] Executing task 'systemctl'
elasticsearch: ACTIVE
----------------------------------------
supervised services
----------------------------------------
[100.64.134.135] Executing task 'status'
Supervisord is not running on grq.
########################################
mozart
########################################
----------------------------------------
third-party services
----------------------------------------
[100.64.134.172] Executing task 'systemctl'
rabbitmq-server: ACTIVE
[100.64.134.172] Executing task 'systemctl'
redis: ACTIVE
[100.64.134.172] Executing task 'systemctl'
elasticsearch: ACTIVE
----------------------------------------
supervised services
----------------------------------------
[100.64.134.172] Executing task 'status'
Supervisord is not running on mozart.
########################################
metrics
########################################
----------------------------------------
third-party services
----------------------------------------
[100.64.134.51] Executing task 'systemctl'
redis: ACTIVE
[100.64.134.51] Executing task 'systemctl'
elasticsearch: ACTIVE
----------------------------------------
supervised services
----------------------------------------
[100.64.134.51] Executing task 'status'
Supervisord is not running on metrics.
########################################
factotum
########################################
----------------------------------------
supervised services
----------------------------------------
[100.64.134.157] Executing task 'status'
Supervisord is not running on factotum.
########################################
ci
########################################
----------------------------------------
third-party services
----------------------------------------
[100.64.134.108] Executing task 'systemctl'
jenkins: ACTIVE
----------------------------------------
supervised services
----------------------------------------
[100.64.134.108] Executing task 'status'
Supervisord is not running on ci.
########################################
verdi
########################################
----------------------------------------
supervised services
----------------------------------------
[100.64.134.108] Executing task 'status'
Supervisord is not running on verdi.

Otherwise if any of the components show the following error, for example for the grq component:

Code Block

########################################
grq
########################################
[100.64.106.214] Executing task 'status'

Fatal error: Needed to prompt for a connection or sudo password (host: 100.64.106.214), but abort-on-prompts was set to True

Aborting.
Needed to prompt for a connection or sudo password (host: 100.64.106.214), but abort-on-prompts was set to True

Fatal error: One or more hosts failed while executing task 'status'

Aborting.
One or more hosts failed while executing task 'status'

then there is an issue with the configuredKEY_FILENAMEonmozartor theauthorized_keysfile under the component's~/.sshdirectory for userOPS_USER. Resolve this issue before continuing on.

Update all HySDS components:
Code Block
sds update all
If you receive any errors, they will need to be addressed.
Start up all HySDS components:
Code Block
sds start all
View status of HySDS components and services:
Code Block
sds status all
During installation, the latest versions of thelightweight-jobscore HySDS package and theverdidocker image was downloaded. Next we import thelightweight-jobspackage:
Code Block
cd ~/mozart/pkgs sds pkg import container-hysds_lightweight-jobs.*.sdspkg.tar
Finally we copy theverdidocker image to the code bucket (CODE_BUCKETas specified duringsds configure). EnsureVERDI_PRIMER_IMAGEurl is consistent:
Code Block
aws s3 cp hysds-verdi-latest.tar.gz s3://<CODE_BUCKET>/hysds-verdi-latest.tar.gz

Next Steps

Now that you have your HySDS cluster configured, continue on toHello World.

To configure Autoscaling groups for your HySDS cluster, continue on toCreate-AWS-Autoscaling-Group-for-Verdi.

To configure a staging area for your HySDS cluster, continue on toCreate-AWS-Resources-for-Staging-Area.

📖 Related Articles:

Filter by label (Content by label)

showLabels	false
max	12
showSpace	false
sort	title	showSpace	false
cql	label = "hysds_setup"

Have Questions? Ask a HySDS Developer:

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

Live Search

placeholder	Search HySDS Wiki

🚀 Page Information:

Was this page useful?

Yes No

Contribution History:

Contributors

mode	list
showLastTime	true
order	update

Find an Error?

Is this document outdated or inaccurate? Please contact the assigned Subject Matter Expert:

Hook Hua

Versions Compared

Old Version 4

New Version 5

Key

Initial Cluster Setup

Next Steps

Yes No

Page Comparison

Versions Compared

Old Version 4

New Version 5

Key

Initial Cluster Setup

Next Steps

Yes No