text
stringlengths 26
1.5k
|
|---|
Checks
The sas-endpoints-ready Check
The sas-endpoints-ready checks all Kubernetes service resources in the SAS Viya
platform namespace to determine if they are ready. When all the services are ready,
this check will pass. If a particular endpoint has no Subset addresses at all (if the
corresponding deployment has zero replicas), it will be marked ready as well. To
disable this check for a particular service, add the sas.com/readiness-check: disabled
label to the Kubernetes Service resource.
The sas-database-ready Check
The sas-database-ready check validates that the database instance required by
SAS is configured properly and is available for connection. There are two parts to
this check:
1Ensure the database is configured properly in Consul.
2Attempt to connect to the database.
If you want to skip the database connection check, set the environment variable SAS_READINESS_CHECK_DATABASE_READY_CONNECT_ENABLED
to false
.
The sas-oauth-provider-ready Check
The sas-oauth-provider-ready ensures that the OAuth provider for the deployment is
available, and that the SAS Viya platform can retrieve OAuth tokens from it. There
are two parts to this check:
1Ensure that the sas-logon-app Kubernetes Endpoints resource exists and is
ready to receive traffic.
2Attempt to retrieve an OAuth token from the SASLogon service.
Usage
|
ready to receive traffic.
2Attempt to retrieve an OAuth token from the SASLogon service.
Usage
To see the results of the latest readiness check:kubectl wait -n name-of-namespace \ --for=condition=ready pod \ --selector="app.kubernetes.io/name=sas-readiness" \
94Chapter 2 / Installation
|
--timeout=1800s
If the deployment is ready, the command has the return code value 0
, and the
following output is displayed:pod/sas-readiness-hash condition met
If the deployment is not ready, the command has the return code value 1
, and the
following output is displayed:pod/sas-readiness-hash condition meterror: timed out waiting for the condition on pod/sas-readiness-hash
Review the Logs
For details of the readiness check’s findings, review its logs:
Note: The command should be entered as a single line.kubectl logs -n name-of-namespace --selector="app.kubernetes.io/name=sas-readiness"
If the deployment is ready, the log will contain a message like the following:"message":"All checks passed. Marking as ready. The first recorded failure was 30s ago."
In order to prevent repetitive log messages, the readiness check only records the
first success message despite continuing to perform its checks. As a result, the log
might appear to contain a stale success message, but no new messages will be
added until the status changes to “not ready”.
The log will contain a message like the following to indicate which services are
failing the readiness check:"message":"The check \"sas-endpoints-ready\" failed - 6 endpoints have no available addresses: sas-catalog,sas-catalog-table-bot,sas-data-flows,sas-device-management,sas-relationships,sas-studio-app"
Customize the Readiness Check Period
By default, the readiness service runs its set of checks every 30 seconds. You can
|
Customize the Readiness Check Period
By default, the readiness service runs its set of checks every 30 seconds. You can
change this value by setting the SAS_READINESS_CHECK_PERIOD
environment
variable on the sas-readiness pod.Readiness Service 95
|
Sign In as the sasboot User
Your SAS environment is deployed with an initial administrator account that is
named sasboot. The password for this account has expired by default, so you must
reset the password before you can sign in.
To reset the password:
1Get the name of the pod that contains SASLogon and search for the characters, sasboot
: kubectl -n name-of-namespace logs -c sas-logon-app $(kubectl -n name-of-namespace get pods -l "app==sas-logon-app" --no-headers --sort-by=.status.startTime -o custom-columns=":metadata.name" | tail -1) | grep "/SASLogon/reset_password?"
Note: This command will not return a value if you set the sas.logon.initial.password
property in a sitedefault.yaml file.
2Sign in from a URL with this format:
protocol ://name-of-ingress-host :port/SASLogon/reset_password?
code=password
protocol can be http or https, depending on how you have secured your
deployment.
3Follow the instructions on the displayed web page to reset the password.
If the URL has expired, restart the sas-logon pod:kubectl -n name-of-namespace delete pods -l "app==sas-logon-app"
Then go to the log and obtain the new URL. The URL expires 24 hours after the
SAS Logon service restarts. For security purposes, the URL that is specified in a
browser or in a text editor also expires, even if the password is not reset.
After you reset the password, SAS Environment Manager automatically opens in
your browser.
|
After you reset the password, SAS Environment Manager automatically opens in
your browser.
4Click Yes for all of the assumable groups so that you have the permissions to
perform subsequent tasks.
5Share the following URL with any other users of your SAS Viya platform so that
they can access the deployment:
protocol ://name-of-ingress-host :port/SASDrive
protocol can be http or https, depending on how you have secured your
deployment.
After the password has been reset, you should consider disabling the password
reset feature. For the steps to disable that feature, see “Disable the Password Reset
Feature” in SAS Viya Platform: Identity Management .96Chapter 2 / Installation
|
Note: If you have enabled multi-tenancy in your deployment, using the sasboot
user to reset the password opens SAS Environment Manager on the provider
tenant.
Promotion and Migration
If you plan to move pre-existing content from SAS 9 or SAS Viya 3, see System
Migration and Content Migration .Promotion and Migration 97
|
98Chapter 2 / Installation
|
3
Post-Installation Tasks
Configure Identities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Configure an LDAP Identity Provider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Configure a SCIM Identity Provider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Configure Multi-tenancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Configure the Connection to the Mail Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Configure Files Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Configure Monitoring and Logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Configure Guest Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Obtain and Run Hadoop Tracer Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Transfer the Hadoop Tracer Script to the Hive Node . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
|
Transfer the Hadoop Tracer Script to the Hive Node . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Tasks Before Running the Hadoop Tracer Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Runtime Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Run the Hadoop Tracer Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Determine If the Run Was Successful . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
How to Address Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Copy the JAR and Configuration Files to the SAS Client System . . . . . . . . . . . . . . . 107
Configure Cloud Analytic Services (CAS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Configure External Access to Amazon Web Services CAS Services . . . . . . . . . . . . 108
Configure Cloud Data Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Configure Model Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Configure Access to Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
|
Configure Access to Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Configure Access to Analytic Store Model Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Configure SAS Asset and Liability Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Configure SAS Business Orchestration Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Configure SAS Data Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Additional Software Associated with SAS Data Engineering . . . . . . . . . . . . . . . . . . . . 112
Configure SAS Data Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Configure the Quality Knowledge Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Configure SAS Dynamic Actuarial Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11399
|
Deploy SAS Enterprise Session Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Configure SAS Event Stream Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Configure SAS Expected Credit Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Configure SAS for Microsoft 365 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Configure SAS Intelligent Decisioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Configure SAS Model Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Configure SAS Model Risk Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Configure SAS Risk Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Configure SAS Risk Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Configure SAS Stress Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Configure SAS Visual Analytics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Additional Software Associated with SAS Visual Analytics . . . . . . . . . . . . . . . . . . . . . . 116
|
Additional Software Associated with SAS Visual Analytics . . . . . . . . . . . . . . . . . . . . . . 116
Configure SAS Visual Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Additional Software Associated with SAS Visual Statistics . . . . . . . . . . . . . . . . . . . . . . 116
Configure SAS Viya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Additional Software Associated with SAS Viya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Configure SAS Viya Advanced . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Additional Software Associated with SAS Viya Advanced . . . . . . . . . . . . . . . . . . . . . . 117
Configure SAS Viya Enterprise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Additional Software Associated with SAS Viya Enterprise . . . . . . . . . . . . . . . . . . . . . . 118
Configure SAS Viya Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Additional Software Associated with SAS Viya Programming . . . . . . . . . . . . . . . . . . . 119
Configure SAS Viya with SingleStore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
|
Configure SAS Viya with SingleStore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Set the Path to the Certificate Authority for SingleStore . . . . . . . . . . . . . . . . . . . . . . . . 120
Additional Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Configure SAS Workload Orchestrator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Configure SAS/CONNECT Spawner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Configure Identities
Configure an LDAP Identity Provider
After completing the installation of the SAS Viya platform, you must configure an
identity provider before your users can access SAS Environment Manager and SAS
Visual Analytics. LDAP is the default identity provider. Make sure your LDAP server
meets the requirements that are described in LDAP Requirements .
For a full set of instructions for configuring an LDAP identity provider, see “How to
Configure LDAP” in SAS Viya Platform: Identity Management .100 Chapter 3 / Post-Installation Tasks
|
Configure a SCIM Identity Provider
After completing the installation of the SAS Viya platform, you must configure the
connection to your identity provider before your users can access SAS Environment
Manager and other SAS Viya platform products. To use SCIM as your source for
user and group identities, you must grant Microsoft Azure access to the SAS Viya
platform SCIM endpoints.
Microsoft Azure
To complete SCIM identity provider configuration, you must configure the SCIM
connection to the SAS Viya platform in Microsoft Azure Active Directory. You also
need to disable LDAP, which is configured by default when the deployment
completes. For a full set of instructions, see “How to Configure SCIM” in SAS Viya
Platform: Identity Management .
Amazon Web Services
You can use Amazon Web Services to access the SAS Viya platform SCIM
endpoints using SCIM as your source for user and group identities. The SAS Viya
platform should support authentication and identity management with Amazon Web
Services Identity and Access Management, with only the initial configuration linking
the SAS Viya platform to AWS IAM. Authentication should be delegated to AWS and
identity information should be synchronized into the SAS Viya platform with the
SCIM identity provider configuration.
Users can change the provisioning from manual to SCIM by following the
instructions to enable automatic provisioning at Enabling Single Sign-On Between
OneLogin and AWS .
Google Cloud Platform
|
instructions to enable automatic provisioning at Enabling Single Sign-On Between
OneLogin and AWS .
Google Cloud Platform
You can utilize Google Cloud Platform to access the SAS Viya platform SCIM
endpoints to use SCIM as your source for user and group identities. To set up
Google Cloud for SCIM, follow the instructions on the web page.
Configure Multi-tenancy
After deployment, a Kubernetes administrator with elevated permissions can
onboard and offboard tenants, monitor services, monitor logs, stop and start Configure Multi-tenancy 101
|
tenants, and run multi-tenant backups and restores. SAS recommends onboarding
at least one tenant into a multi-tenant environment immediately after deployment.
For information about performing each of these tasks, see SAS Viya Platform: Multi-
tenancy .
Configure the Connection to the Mail
Service
After performing a new deployment of the SAS Viya platform, you must configure
the connection to your mail service. Complete these steps while you are signed in
as one of the SAS Administrators.
1Select the from the side menu to open the Configuration page.
2On the Configuration page, select All Services from the list, and then select
Mail service from the list of services.
Note: If Mail service is not listed, then there is no configuration to be
performed. You should skip the rest of this topic.
3In the sas.mail section, click . In the Edit Configuration window, follow these
steps:
aSpecify a value for the following required fields: host , port, and
fromAddress . For the remaining fields, review the default values and make
changes, as necessary. The default values are appropriate for most sites.
bClick Save .
4(Optional) To enable the health check for the mail service, perform the following
steps.
aSelect the from the side menu to open the Configuration page.
bOn the Configuration page, select All Services from the list, and then select
Mail service from the list of services.
cIn the management.health.mail section, click .
dTurn the enabled toggle to on
|
Mail service from the list of services.
cIn the management.health.mail section, click .
dTurn the enabled toggle to on
.
eClick Save .
When this toggle is set, health checks will be enabled after the mail service is
restarted. If the mail host is not configured or is configured incorrectly, or if it cannot
connect to the SMTP mail server, the mail service will indicate it is in a failed state.102 Chapter 3 / Post-Installation Tasks
|
Configure Files Service
After deploying the SAS Viya platform, the Files service must be configured to
provide in-line responses to file contents. For the steps to perform this configuration,
see “Managing Cross-Site Scripting Risk for File Uploads” in SAS Viya Platform:
Overview .
Configure Monitoring and Logging
Monitoring performance metrics and the logs generated by your deployment helps
your SAS administrators ensure that the SAS Viya platform is running efficiently and
enable them to quickly detect any problems. You can use your preferred monitoring
technology to monitor your SAS Viya platform deployment. If you do not have a
preferred technology, SAS provides solutions for monitoring and logging that are
based on widely-used open source monitoring technologies including Prometheus,
Grafana, and OpenSearch.
You can deploy these solutions from the SAS Viya Platform Monitoring for
Kubernetes GitHub site.
The monitoring deployment includes these components:
nPrometheus
nPrometheus Operator
nAlert Manager
nGrafana
nPrometheus Pushgateway
nGrafana dashboards
The logging deployment includes these components:
nFluent Bit
nOpenSearch
nOpenSearch Dashboards
Configure Guest Access
Guest access is an optional feature that provides anonymous Read-Only access to
a subset of resources and functionality in participating applications. For information Configure Guest Access 103
|
about configuring your SAS Viya platform deployment for guest access, see “Guest
Access ” in SAS Viya Platform: Authentication .
Obtain and Run Hadoop Tracer Script
Note: The Hadoop tracer script is used with SAS/ACCESS Interface to Hadoop and
SAS/ACCESS Interface to Impala. The Hadoop tracer script might also be required
by additional components that access Hadoop. If you do not need to run the
Hadoop tracer script, then skip this section.
Overview
The Hadoop tracer script is a Python script that traces the system calls of various
Hadoop client tools and uses the trace data to identify required client JAR files and
configuration files. The script determines which specific Hadoop distribution supplied
the JARs and configuration files that SAS products need for connectivity between
Hadoop client machines and the Hadoop server environment. Therefore, you must
run this script on the Hive node of your Hadoop cluster in order to prepare your
Hadoop environment for a SAS Viya platform deployment.
The script is organized into two parts. The first part is a list of the Hadoop services
that the script is tracing to a new driver.json file. The second part is the remainder of
the code. This division allows the list of Hadoop services and additional JAR files to
be modified manually without having to make Python code changes.
If the driver.json file is moved or removed from its default location, the Hadoop
|
be modified manually without having to make Python code changes.
If the driver.json file is moved or removed from its default location, the Hadoop
tracer script fails with an error message indicating that the file is missing.
Transfer the Hadoop Tracer Script to the Hive Node
The Hadoop tracer script runs on the Hive node of your Hadoop cluster.
1On your Hadoop cluster, create a temporary directory to hold a ZIP file that you
download later. For example, /tmp/sas/hadooptracer
.
2Copy and paste the following URL into a browser. Download the
hadooptracer.zip file from that FTP site to the directory that you created in step
1.
ftp.sas.com/techsup/download/blind/access/hadooptracer.zip
3Using a method of your choice (such as PSFTP, SFTP, SCP, or FTP), transfer
the ZIP file to the Hive node on your Hadoop cluster.
4Unzip the file. The hadooptracer_py and driver.json files are included in the ZIP
file.104 Chapter 3 / Post-Installation Tasks
|
5Change permissions on the hadooptracer_py file to include the Execute
permission:chmod 755 ./hadooptracer_py
Tasks Before Running the Hadoop Tracer Script
nEnsure that the user running the script has authorization to issue HDFS and Hive
commands.
nIf Hadoop is secured with Kerberos, obtain a Kerberos ticket for the user before
running the script.
nEnsure that Python and the strace Linux library have been installed on the
Hadoop cluster. Install them from the package repositories for your Linux
distribution if necessary. Python 2.6 or later is required.
nIf you want to pull one or more JAR files that are not included in the output of the
hadoop tracer script, modify the driver.json file by adding the JAR files to the
ExtraJarFiles property at the bottom of the file. Here is an example of adding a
new JAR file named xyz-service.jar to the ExtraJarFiles property:"ExtraJarFiles": ["jline-*","jruby-complete-*","hive-warehouse-connector-assembly-*","xyz-service-*"]
Runtime Options
For a list of the options available when running the Hadoop tracer script:python ~/hadooptracer_py --help
Review these options before you run the script.
Run the Hadoop Tracer Script
Use this command to run the Hadoop tracer script:python ./hadooptracer_py --filterby=latest --postprocess --jsonfile ./driver.json
If you want to pull the necessary JAR files without filtering, use filterby=none
, or
omit the filterby=
option.
|
If you want to pull the necessary JAR files without filtering, use filterby=none
, or
omit the filterby=
option.
TIP To collect only the Hadoop configuration files and exclude JAR files:python ./hadooptracer_py --svckey=Hadoop
Obtain and Run Hadoop Tracer Script 105
|
TIP The postprocess option ensures that the ${hdp.version}
tokens are
replaced. SAS strongly recommends that you run the tracer script with this
option. This option is ignored for Cloudera clusters.
The Hadoop tracer script performs the following tasks:
ncollects the necessary Hadoop JAR and configuration files from nodes in the
cluster and copies them in the /tmp/jars
directory and the /tmp/sitexmls
directory, respectively.
ncreates a hadooptracer.json file in the /tmp directory. If you need a custom path
for the JSON output file, use this command instead:python ./hadooptracer_py –f /your-path/hadooptracer.json
ncreates a log at /tmp/hadooptracer.log
. If you need a custom path for the log
file, add this option:--logfile your-path/log-file-name
nprints more debug entries to the log file when the --debug
option is added.
IMPORTANT The Hadoop JAR and configuration files on the SAS client
machine must be kept in sync with the Hadoop configuration. After a Hadoop
cluster update, re-run the Hadoop tracer script to collect any new files, and
then copy those files to the SAS client machine, replacing any previous
versions of the files.
Determine If the Run Was Successful
1Ensure that the required Hadoop JAR files are collected from the Hadoop cluster
and placed in the ./jars
directory.ls -l ./jars
2Ensure that the required Hadoop configuration files are collected from the
Hadoop cluster and placed in ./confs
directory.ls -l ./confs
|
2Ensure that the required Hadoop configuration files are collected from the
Hadoop cluster and placed in ./confs
directory.ls -l ./confs
3Review the hadooptracer.log file that is located in the default location, /tmp, or
the custom location that you specified.
How to Address Failures
Most errors with the Hadoop tracer script stem from improper usage or an incorrect
cluster configuration. If there are a problems with the Hadoop cluster, they will
typically show up in the stdout of the Hadoop tracer script in the form of Java
traceback information.106 Chapter 3 / Post-Installation Tasks
|
Another common problem occurs when users try to run the Hadoop tracer script on
a cluster node that doesn't have Hadoop/Hive/HDFS/Yarn/Pig/etc
in an available
PATH. For example,2020-04-07 12:16:51,036 hadooptracer [ERROR] pig is not found in the $PATH
Inspect hadooptracer.log, located in /tmp by default, and use the rest of this
troubleshooting section to resolve common issues. Some error messages in the
console output for hadooptracer_py are normal and do not necessarily indicate a
problem with the JAR and configuration file collection process. However, if the files
are not collected as expected or if you experience problems connecting to Hadoop
with the collected files, contact SAS Technical Support and include the
hadooptracer.log and the hadooptracer.json files.
Copy the JAR and Configuration Files to the SAS
Client System
SAS 9.4 and SAS Viya 3.5
1On the SAS client machine, create two directories to hold the JAR and
configuration files. For example, the /opt/sas/hadoopfiles/lib
and /opt/sas/hadoopfiles/conf
directories.
2Using a method of your choice (such as PSFTP, SFTP, SCP, or FTP), copy the
files in the tmp/jars
and tmp/confs
directories on the Hadoop server to the
directories on the SAS client machine that you created.
Note: If you connect to the Hadoop server with an HTTP REST API, you do not
need the Hadoop JAR files on the SAS client machine.
The SAS Viya Platform in a Kubernetes
Environment
|
need the Hadoop JAR files on the SAS client machine.
The SAS Viya Platform in a Kubernetes
Environment
For more information about configuring SAS/ACCESS Interface to Hadoop, see the
README file at $deploy/sas-bases/examples/data-access/README.md
(for
Markdown) or at $deploy/sas-bases/docs/configuring_sasaccess_and_data_connectors_for_sas_viya_4.htm
(for HTML).Obtain and Run Hadoop Tracer Script 107
|
Configure Cloud Analytic Services (CAS)
Configure External Access to Amazon Web
Services CAS Services
To gain secure external access to CAS services running in an AWS cloud, additional
steps are necessary after the load balancer setup described in “Configure External
Access to CAS” on page 58 has been completed and the cluster has been
deployed.
All load balancer hosts must be DNS-aliased to the service names generated by
setting up the load balancer. You may select an extant DNS domain/subdomain or
create a new one specifically to support access to AWS CAS clusters. You may use
your own DNS service or transfer DNS management of the domain to the AWS
Route 53 service.
Note: AWS Route 53 offers some advantages for large scale or more complex
deployment cases due to the additional routing and monitoring capabilities that it
offers. See What is Amazon Route 53? for details.
Regardless of the DNS service used, you must add CNAME records to the routing
table to direct references to the CAS-generated names to the load balancer FQDN.
CAS-generated external service names are predetermined based upon the suffix
supplied in the earlier setup. Here are examples of those names when using Data
|
supplied in the earlier setup. Here are examples of those names when using Data
Connect Accelerators and Spark EPCS access in addition to the general ports:sas-cas-server-default-bin-pisces.viya.acme.comsas-cas-server-default-epcs-pisces.viya.acme.comsas-cas-server-default-controller-dc-pisces.viya.acme.comsas-cas-server-default-backup-dc-pisces.viya.acme.comsas-cas-server-default-worker-0-dc-pisces.viya.acme.comsas-cas-server-default-worker-1-dc-pisces.viya.acme.com
To find the load balancer names:kubectl -n name-of-namespace get svc | grep LoadBalancer
Here is a partial sample of typical output, revealing the load balancer DNS names:108 Chapter 3 / Post-Installation Tasks
|
sas-cas-server-default-backup-dc LoadBalancer 10.100.178.234 a7cc66edbca1d402c8c72ea7da3543d7-1688256446.us-east-1.elb.amazonaws.comsas-cas-server-default-bin LoadBalancer 10.100.169.27 a5a90408edfc246bba051f830d6e58fe-2016925591.us-east-1.elb.amazonaws.comsas-cas-server-default-controller-dc LoadBalancer 10.100.30.89 a14ca0a2c8009415891c05053ee36d83-222471472.us-east-1.elb.amazonaws.comsas-cas-server-default-epcs LoadBalancer 10.100.105.198 a8207fb65f9464e7680948e8e8d9e722-58303842.us-east-1.elb.amazonaws.comsas-cas-server-default-http LoadBalancer 10.100.65.146 aa21afd024de44d64bffe557add1e47a-1604903159.us-east-1.elb.amazonaws.comsas-cas-server-default-worker-0-dc LoadBalancer 10.100.4.214 a307341cce8694e639c6fdee9096cb5c-632571573.us-east-1.elb.amazonaws.comsas-cas-server-default-worker-1-dc LoadBalancer 10.100.224.50 a04b03687b3704b79820fb72869eb5c5-1997141149.us-east-1.elb.amazonaws.com
With this information you can then cut your CNAME records in your DNS service of
choice in this arrangement:sas-cas-server-default-backup-dc-pisces.viya.acme.com ⇒ a7cc66edbca1d402c8c72ea7da3543d7-1688256446.us-east-1.elb.amazonaws.comsas-cas-server-default-bin-pisces.viya.acme.com ⇒ a5a90408edfc246bba051f830d6e58fe-2016925591.us-east-1.elb.amazonaws.comsas-cas-server-default-controller-dc-pisces.viya.acme.com ⇒ a14ca0a2c8009415891c05053ee36d83-222471472.us-east-1.elb.amazonaws.com...
Configure Cloud Data Exchange
|
Configure Cloud Data Exchange
Note: For a full description of the deployment of Cloud Data Exchange and SAS
Data Agent, see “What is the deployment process for SAS Data Agent?” in Getting
Started with SAS Viya Platform Operations .
Cloud Data Exchange is included in a number of offerings from SAS:
nSAS Data Engineering Advanced
nSAS Intelligent Decisioning
nSAS Visual Analytics
nSAS Visual Forecasting
nSAS Visual Statistics
nSAS Visual Text Analytics
nSAS Viya
nSAS Viya Advanced
nSAS Viya Enterprise
nSAS Viya Programming
If you have at least one of these offerings in your software order, you must complete
your deployment of Cloud Data Exchange to take advantage of its data
management and connection capabilities.Configure Cloud Data Exchange 109
|
After you have finished your deployment of the SAS Viya platform, you must
perform a few more tasks to finish deploying Cloud Data Exchange:
1Ensure that you have deployed the co-located SAS Data Agent by following the
steps in the README file located at $deploy/sas-bases/examples/sas-data-agent-server-colocated/README.md
(for Markdown format) or at $deploy/sas-bases/docs/configure_a_co-located_sas_data_agent.htm
(for HTML
format).
2Deploy the remote SAS Data Agent using the procedures described in Remote
SAS Data Agent: Deployment Guide .
3Perform the post-installation tasks described at “Once Co-located SAS Data
Agent and Remote SAS Data Agent Are Running” in Cloud Data Exchange for
the SAS Viya Platform: Administrator’s Guide .
Configure Model Access
Configure Access to Models
Note: SAS Model Manager is only available as part of the SAS Model Manager,
SAS Viya, SAS Viya Advanced, and SAS Viya Enterprise offerings.
In order to import models into SAS Model Manager and register models from Model
Studio, SAS Visual Analytics, and SAS Studio into the common model repository, as
well as add a model from the common model repository into a decision flow, users
must have the appropriate access permissions. For more information, see “Access
to Models” in SAS Viya Platform: Models Administration .
Configure Access to Analytic Store Model Files
Note: SAS Model Manager is only available as part of the SAS Model Manager,
|
Configure Access to Analytic Store Model Files
Note: SAS Model Manager is only available as part of the SAS Model Manager,
SAS Viya, SAS Viya Advanced, and SAS Viya Enterprise offerings. SAS Intelligent
Decisioning is only available as part of the SAS Viya Enterprise and SAS Intelligent
Decisioning offerings. If your deployment does not include any of these offerings,
you should skip this section.
CAUTION
To use analytic store models, you must have completed the tasks described in the
README file, ”SAS Micro Analytic Service ASTORE Configuration”. The README file
islocated at $deploy/sas-bases/examples/sas-microanalytic-score/astores/
110 Chapter 3 / Post-Installation Tasks
|
README.md
(for Markdown) and $deploy/sas-bases/docs/sas_micro_analytic_service_astore_configuration.htm
(for HTML).
In order to publish analytic store models from SAS Model Manager, Model Studio,
and SAS Intelligent Decisioning to the SAS Micro Analytic Service destination, see
“Accessing Analytic Store Model Files” in SAS Viya Platform: Models Administration .
Configure SAS Asset and Liability
Management
To complete the configuration of this product, you must perform the post-installation
tasks described in SAS Asset and Liability Management: Administrator’s Guide
(access key required).
Configure SAS Business Orchestration
Services
SAS Business Orchestration Services requires a set of configuration files. When you
have completed the deployment of the SAS Viya platform and SAS Business
Orchestration Services, you must create an init container image that includes all of
the required files, and you must add a reference to the init container to the base
kustomization.yaml file. Instructions are provided in a README file located at $deploy/sas-bases/examples/sas-boss/README.md
(for Markdown format) or $deploy/sas-bases/docs/deploying_sas_business_orchestration_services.htm
(for HTML format).
The product documentation provides additional instructions for configuring and
using SAS Business Orchestration Services. The documentation is available from
the SAS Support website at https://support.sas.com/en/software/business-
|
the SAS Support website at https://support.sas.com/en/software/business-
orchestration-support.html .Configure SAS Business Orchestration Services 111
|
Configure SAS Data Engineering
Additional Software Associated with SAS Data
Engineering
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for Python
nSWAT for R
A SAS Viya 3.5 software order is associated with deploying SAS Data Engineering.
The software order contains SAS Embedded Process for Hadoop. The Software
Order Email (SOE) specifies the location of the deployment documentation: SAS
Viya In-Database Technologies: Deployment and Administration Guide . You should
deploy this order only if you are using SAS In-Database Technologies for Hadoop
Cloud Services.
A second SAS Viya platform order is also associated with deploying SAS Data
Engineering. It contains SAS Embedded Process for Teradata and SAS Embedded
Process for Spark. The Software Order Email (SOE) specifies the location of the
deployment documentation: SAS Viya In-Database Technologies: Deployment and
Administration Guide . You should deploy this order only if you are using SAS In-
Database Technologies for Teradata or SAS In-Database Technologies for Spark.
Note: The SOEs for the associated orders do not refer to SAS Data Engineering
explicitly.
Configure SAS Data Quality
Configure the Quality Knowledge Base
For an overview of SAS Data Quality, see “Overview” in SAS Viya Platform: QKB
|
Configure SAS Data Quality
Configure the Quality Knowledge Base
For an overview of SAS Data Quality, see “Overview” in SAS Viya Platform: QKB
Management . SAS Data Quality relies on a collection of rules and reference data
called a SAS Quality Knowledge Base (QKB).
Your deployment includes the latest version of the QKB for Contact Information. It is
configured to use the English, United States
locale by default. To deploy a 112 Chapter 3 / Post-Installation Tasks
|
custom QKB into CAS, you should first create a QKB Archive and import it into CAS
using SAS Environment Manager. For information about these tasks, see “Create a
QKB Archive (QARC) File” in SAS Viya Platform: QKB Management .
If you are upgrading from SAS Viya 3.5, you must manually import any older or
custom QKBs into this version. For instructions to import the QKB into SAS Studio,
see “How To (Kubernetes)” in SAS Viya Platform: QKB Management and “Set the
Default QKB and the Default Locale” in SAS Viya Platform: QKB Management . To
import a QKB into CAS, create a QARC archive and use the SAS Environment
Manager; for details, see “Create a QKB Archive (QARC) File” in SAS Viya
Platform: QKB Management .
Configure SAS Dynamic Actuarial
Modeling
Additional tasks are required to deploy this product. For details, see SAS Dynamic
Actuarial Modeling: Administrator’s Guide (access key required).
Deploy SAS Enterprise Session Monitor
SAS Enterprise Session Monitor extends the monitoring and scheduling capabilities
of your SAS Viya platform deployment by providing users with easy access to
metrics that lead to resource optimization and enhanced throughput. However,
deploying SAS Enterprise Session Monitor requires a separate download and
deployment process from the rest of the SAS Viya platform.
If you have ordered SAS Enterprise Session Monitor, you will have received an
email, separate from your Software Order Email, describing where and how to
|
email, separate from your Software Order Email, describing where and how to
download the required software. The email describes the location of the
deployment, administration, and usage documentation. Follow the instructions in
that email to download and deploy SAS Enterprise Session Monitor.
Configure SAS Event Stream Processing
In order to start using SAS Event Stream Processing, you must first use SAS Event
Stream Processing Studio to launch a project. This causes the ESP Operator to
start an ESP Server pod, which then runs the project.
For more information, see “Running Event Stream Processing Projects” in SAS
Event Stream Processing: Overview .Configure SAS Event Stream Processing 113
|
Configure SAS Expected Credit Loss
Additional tasks are required to deploy this product. For details, see SAS Expected
Credit Loss: Administrator’s Guide (access key required).
Configure SAS for Microsoft 365
For the information to complete the configuration of this product, including required
security configuration, see “Steps for SAS Administrator” in SAS for Microsoft 365:
User’s Guide .
Configure SAS Intelligent Decisioning
Note: SAS Intelligent Decisioning is only available as part of the SAS Viya
Enterprise and SAS Intelligent Decisioning offerings. If your deployment does not
include any of these offerings, you should skip this section.
You must perform post-installation tasks such as configuring access to analytic store
model files and setting configuration properties for SAS Intelligent Decisioning. For
more information, see SAS Intelligent Decisioning: Administrator’s Guide .
Configure SAS Model Manager
Note: SAS Model Manager is only available as part of the SAS Model Manager,
SAS Viya, SAS Viya Advanced, and SAS Viya Enterprise offerings. If your
deployment does not include any of these offerings, you should skip this section.
You must perform post-installation tasks such as configuring access to analytic store
model files and a workflow client user account for use with SAS Model Manager. For
more information, see SAS Model Manager: Administrator’s Guide .114 Chapter 3 / Post-Installation Tasks
|
Configure SAS Model Risk Management
To complete the configuration of this product, you must perform post-installation
tasks to configure additional access control. For more information, see “Designing
and Implementing Access Control” in SAS Risk Cirrus: Administrator’s Guide .
Configure SAS Risk Engine
To complete the configuration of SAS Risk Engine, you must perform post-
installation tasks such as verifying the deployment and granting users access to
objects in the user interface. For more information, see SAS Risk Engine:
Administration (access key required).
Configure SAS Risk Modeling
To complete the configuration of this product, you must perform post-installation
tasks such as working with user groups and creating a destination for publishing a
model. For more information, see SAS Risk Modeling: Administrator’s Guide
(access key required).
Configure SAS Stress Testing
Additional tasks are required to deploy this product. For details, see SAS Stress
Testing: Administrator’s Guide (access key required).Configure SAS Stress Testing 115
|
Configure SAS Visual Analytics
Additional Software Associated with SAS Visual
Analytics
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for Python
nSWAT for R
A SAS 9.4 order is associated with deploying SAS Visual Analytics. The software
order contains SAS Visual Analytics Add-In for Office. The SOE specifies the
location of the deployment documentation.
Note: The SOE for the associated order does not refer to SAS Visual Analytics
explicitly.
Configure SAS Visual Statistics
Additional Software Associated with SAS Visual
Statistics
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for Python
nSWAT for R
A SAS 9.4 order is associated with deploying SAS Visual Statistics. The software
order contains SAS Visual Analytics Add-In for Office. The SOE specifies the
location of the deployment documentation.116 Chapter 3 / Post-Installation Tasks
|
Note: The SOE for the associated order does not refer to SAS Visual Statistics
explicitly.
Configure SAS Viya
Additional Software Associated with SAS Viya
Note: Before the 2023.01 version of the SAS Viya platform, this offering was
named SAS Visual Machine Learning.
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for Python
nSWAT for R
A SAS 9.4 order is associated with deploying SAS Viya. The software order
contains SAS Visual Analytics Add-In for Office. The SOE specifies the location of
the deployment documentation.
Note: The SOE for the associated order does not refer to SAS Viya explicitly.
Configure SAS Viya Advanced
Additional Software Associated with SAS Viya
Advanced
Note: Before the 2023.01 version of the SAS Viya platform, this offering was
named SAS Visual Data Science.
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for PythonConfigure SAS Viya Advanced 117
|
nSWAT for R
A SAS 9.4 order is associated with deploying SAS Viya Advanced. The software
order contains SAS Visual Analytics Add-In for Office. The SOE specifies the
location of the deployment documentation.
Note: The SOE for the associated order does not refer to SAS Viya Advanced
explicitly.
Configure SAS Viya Enterprise
Additional Software Associated with SAS Viya
Enterprise
Note: Before the 2023.01 version of the SAS Viya platform, this offering was
named SAS Visual Data Science Decisioning.
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for Python
nSWAT for R
A SAS Viya 3.5 software order is associated with deploying SAS Viya Enterprise.
The software order contains SAS Embedded Process for Hadoop. The Software
Order Email (SOE) specifies the location of the deployment documentation: SAS
Viya In-Database Technologies: Deployment and Administration Guide . You should
deploy this order only if you are using SAS In-Database Technologies for Hadoop
Cloud Services.
A SAS 9.4 order is associated with deploying SAS Viya Enterprise. The software
order contains SAS Visual Analytics Add-In for Office. The SOE specifies the
location of the deployment documentation.
A second SAS Viya platform order is also associated with deploying SAS Viya
Enterprise. It contains SAS Embedded Process for Teradata and SAS Embedded
|
Enterprise. It contains SAS Embedded Process for Teradata and SAS Embedded
Process for Spark. The Software Order Email (SOE) specifies the location of the
deployment documentation: SAS Viya In-Database Technologies: Deployment and
Administration Guide . You should deploy this order only if you are using SAS In-
Database Technologies for Teradata or SAS In-Database Technologies for Spark.
Note: The SOEs for the associated orders do not refer to SAS Viya Enterprise
explicitly.118 Chapter 3 / Post-Installation Tasks
|
Configure SAS Viya Programming
Additional Software Associated with SAS Viya
Programming
Note: Before the 2023.01 version of the SAS Viya platform, this offering was
named SAS Data Science Programming.
If you want to access Cloud Analytic Services (CAS) with R or Python, you must
download and deploy SWAT packages for them. Those packages are available at
the following locations:
nSWAT for Python
nSWAT for R
A SAS Viya 3.5 software order is associated with deploying SAS Viya Programming.
The software order contains SAS Embedded Process for Hadoop and SAS
Embedded Process for Teradata on Linux. The Software Order Email (SOE)
specifies the location of the deployment documentation: SAS Viya In-Database
Technologies: Deployment and Administration Guide . You should deploy this order
only if you are using SAS In-Database Technologies for Hadoop Cloud Services or
SAS In-Database Technologies for Teradata.
A SAS 9.4 order is associated with deploying SAS Viya Programming. The software
order contains SAS Visual Analytics Add-In for Office. The SOE specifies the
location of the deployment documentation.
A second SAS Viya platform order is also associated with deploying SAS Viya
Programming. It contains SAS Embedded Process for Spark.
Note: The SOEs for the associated orders do not refer to SAS Viya Programming
explicitly.Configure SAS Viya Programming 119
|
Configure SAS Viya with SingleStore
Set the Path to the Certificate Authority for
SingleStore
Encryption between CAS and SingleStore is supported under the same TLS mode
that you set for SAS Viya. However, you can specify the path to a different certificate
authority file by setting the ssl_ca=
option in the SingleStore data connector. For
more information about the ssl_ca=
option, see “Security” in SAS Viya Platform with
SingleStore: Administration and Configuration Guide .
Additional Documentation
For more information about configuring and administering SAS Viya with
SingleStore, see SAS Viya Platform with SingleStore: Administration and
Configuration Guide .
Configure SAS Workload Orchestrator
Kubernetes provides capabilities to match workload to compute nodes so that jobs
can be scheduled and executed. It determines which nodes can be used to start
pods for work, based on available resources. SAS Workload Orchestrator extends
the workload management capabilities of Kubernetes by adding priority-based
queues, which provide the ability to control where and when jobs are processed
based on prioritization and resource allocation.
Beginning in version 2023.08, SAS Workload Orchestrator is enabled by default with
your SAS Viya platform deployment. For more information about using SAS
Workload Orchestrator, including how to disable it, see SAS Viya Platform:
Workload Management .120 Chapter 3 / Post-Installation Tasks
|
Configure SAS/CONNECT Spawner
If you are performing a new deployment and added a reference to the enable-
spawned-servers.yaml example file to your base kustomization.yaml in order to
spawn SAS/CONNECT servers in the spawner pod, you must make a change to the
configuration in SAS Environment Manager.
1Go to SAS Environment Manager.
2Select Configurations , then SAS/CONNECT Spawner .
3In the contents field for sas.connect.spawner: startup_commands , remove -nolocallaunch
from the USERMODS line.
4Click Save .
For more information about this task, see the “Allow the Ability to Spawn Servers
within the Spawner Pod” section of the “Configure SAS/CONNECT Spawner in SAS
Viya” README file, located at $deploy/sas-bases/examples/sas-connect-spawner/README.md
(for Markdown format) and at $deploy/sas-bases/docs/configure_sasconnect_spawner_in_the_sas_viya_platform.htm
(for HTML
format).Configure SAS/CONNECT Spawner 121
|
122 Chapter 3 / Post-Installation Tasks
|
4
Validating the Deployment
The SAS Viya Platform: Deployment Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
The SAS Viya Platform and the SAS Operational Quality Tool . . . . . . . . . . . . . . . . . . 123
The SAS Viya Platform: Deployment
Validation
Validating your SAS Viya platform deployment is not a Kubernetes-based task and
requires a SAS administrator. The procedures for validating products in your
deployment are described in SAS Viya Platform: Deployment Validation .
The SAS Viya Platform and the SAS
Operational Quality Tool
SAS Operational Qualification Tool (SAS OQ) helps qualify the use of SAS software
in regulated industries. SAS OQ supports the qualification aspect of the essential
migration, integration, and verification processes that customers must perform to
validate the SAS deployment. SAS OQ helps to demonstrate that SAS is
operational. For more information about SAS OQ, see SAS Operational
Qualification Tool: User’s Guide .123
|
124 Chapter 4 / Validating the Deployment
|
5
Modifying the Deployment
Configuration
Modify Existing Customizations in a Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Modify Existing Customizations in a
Deployment
After your software has been deployed, you might want to modify some of the
customizations that you made in your initial deployment. For example, you might
want to move from an SMP deployment of CAS to an MPP deployment. Or you
might want to move from using TLS in front-door mode to using full-stack TLS.
Modifying your configuration does not require new software. To modify the
customizations, simply repeat some of the steps that you performed in the initial
deployment of your software:
IMPORTANT Performing these steps causes an outage while the software
is re-deployed with the new configuration settings. Ensure that you plan for
an outage before continuing with the steps..
1Perform the modifications. Refer to the topic in “Common Customizations” on
page 36 or the README file that describes the changes that should be
made.The README indexes, located at $deploy/sas-bases/README.md
for
README files in Markdown language or $deploy/sas-bases/docs/index.htm
for README files in HTML, can help direct you to the appropriate README.
2Redeploy the software.
nIf you are deploying with the SAS Viya Platform Deployment Operator, first
modify the SASDeployment custom resource with the command provided at 125
|
“Run the create sas-deployment-cr Command” on page 79. Then run the
deployment command described at “Command and Output” on page 85.
nIf you are deploying with the Orchestration Tool, perform the command
described at “Command” on page 88.
nIf you are deploying with Kubernetes commands, run the series of commands
that are described at “Deployment Using Kubernetes Commands” on page
91.126 Chapter 5 / Modifying the Deployment Configuration
|
6
Uninstalling
Uninstall with the SAS Viya Platform Deployment Operator . . . . . . . . . . . . . . . . . . . . 127
Uninstall a SAS Viya Platform Deployment from a Namespace Only . . . . . . . . . . . . 127
Remove All SAS Viya Platform Deployments in the Cluster and the
SAS Viya Platform Deployment Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Remove the SAS Viya Platform Deployment Operator Only . . . . . . . . . . . . . . . . . . . . 133
Uninstall Deployments That Do Not Use the Deployment Operator . . . . . . . . . . . . . 133
Remove the SAS Prepull and SAS Workload Orchestrator
ClusterRoles and ClusterRoleBindings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Pause the SingleStore Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Remove Service Account Links and SCCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Remove Internal PostgreSQL Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Remove the SAS Viya Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Remove CRDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Uninstall with the SAS Viya Platform
Deployment Operator
|
Uninstall with the SAS Viya Platform
Deployment Operator
The following steps assume that the SAS Viya platform is running and functional. If
the software is not running, start it. If the software is not functioning, contact SAS
Technical Support.
Uninstall a SAS Viya Platform Deployment from a
Namespace Only
The steps in this section remove the SAS Viya platform from a single namespace,
but leave the SAS Viya Platform Deployment Operator running in the cluster. As
long as any namespace in the cluster is running and managed by the operator, do
not remove the operator. If you want to remove all of the deployments from all of the 127
|
namespaces in a cluster and also remove the operator from that cluster, see
“Remove All SAS Viya Platform Deployments in the Cluster and the SAS Viya
Platform Deployment Operator” on page 130.
Remove the SAS Prepull and SAS
Workload Orchestrator ClusterRoles and
ClusterRoleBindings
Before you begin to uninstall your SAS Viya platform software, perform the following
tasks to account for some special cluster-wide permissions.
1If you enabled the Image Staging Node List Option in your deployment, follow
the instructions in the “Disable the Node List Option” section of the “SAS Image
Staging Configuration Options” README located at $deploy/sas-bases/examples/sas-prepull/README.md
(for Markdown format) and $deploy/sas-bases/docs/sas_image_staging_configuration_options.htm
(for HTML
format).
2If your deployment includes Workload Management and Cluster Role was
enabled for it, follow the instructions in the “Disable the Cluster Role” section of
the ”Cluster Privileges for SAS Workload Orchestrator Service” README
located at $deploy/sas-bases/examples/sas-workload-orchestrator/configure/README.md
(for Markdown format) and $deploy/sas-bases/docs/configuration_settings_for_sas_workload_orchestrator_service.htm
(for
HTML format).
Pause the SingleStore Operator
If your deployment includes SAS Viya with SingleStore, you must pause the
SingleStore operator before uninstalling the SAS Viya platform software.
|
If your deployment includes SAS Viya with SingleStore, you must pause the
SingleStore operator before uninstalling the SAS Viya platform software.
1Pause the SingleStore cluster:kubectl -n name-of-namespace patch MemsqlCluster \ sas-singlestore-cluster --type=json -p \ '[{"op": "add", "path": "/spec/clusterState","value": "pause"}]
2When you pause the cluster, each leaf shuts down after snapshots are taken of
all the databases. The snapshot process can lengthen the shutdown time,
especially if a large amount of new data requires a snapshot. To check the status
of the snapshot process, view the log of the SingleStore pods. Here are some
examples:kubectl logs node-sas-singlestore-cluster-master-0 node \-n name-of-namespacekubectl logs node-sas-singlestore-cluster-leaf-ag1-0 node \-n name-of-namespace
128 Chapter 6 / Uninstalling
|
During the snapshots process, the snapshot status is reported as a percent (%)
complete. As shutdown completes, each leaf shows Completed Detaching
in
the log.
3When all of the pods terminate, proceed with uninstalling the SAS Viya platform.
Remove SAS Viya Platform from a
Namespace
When the operator deploys the SAS Viya platform, it assigns an ownerReference
from the custom resource to every namespace-scoped resource deployed by the
operator. When the custom resource is deleted, Kubernetes deletes anything that
was owned by that resource, thereby uninstalling the SAS Viya platform. However,
some resources in a SAS Viya platform deployment are cluster-scoped (such as
CRDs, ClusterRoles, and ClusterRoleBindings). These are shared by all SAS Viya
platform deployments in the cluster, so they are not considered “owned” by any one
custom resource. Therefore, they are not deleted when the custom resource is
removed for a namespace.
1Remove the resources for an internal instance of PostgreSQL from the
namespace:
Note: If your deployment does not include an internal instance of PostgreSQL,
skip this step.kubectl -n name-of-namespace delete postgresclusters --selector="sas.com/deployment=sas-viya"
Note: If your Crunchy Data PVs do not have a reclaimPolicy of "retain",
performing this command will delete their associated PVCs, resulting in the data
they contain being lost. Therefore, ensure that you have backed up that data or
|
they contain being lost. Therefore, ensure that you have backed up that data or
are otherwise unconcerned about losing it.
If your Crunchy Data PVs do have a reclaimPolicy of "retain", their PVCs will not
be deleted when the cluster is deleted. If you no longer need them, you can
delete them manually by using the kubectl get pvc
command to get the list of
PVCs and the kubectl delete
command to delete them.
2Remove the SASDeployment custom resource from the namespace. If you kept
your $deploy -sasdeployment.yaml file, run the following command:kubectl -n name-of-namespace delete -f $deploy-sasdeployment.yaml
If you did not keep the $deploy -sasdeployment.yaml file, run the following
command:kubectl -n name-of-namespace delete sasdeployments SASdeployment-custom-resource-name
Uninstall with the SAS Viya Platform Deployment Operator 129
|
Remove Service Account Links
Note: If your deployment is not on Red Hat OpenShift, skip this section.
For Red Hat OpenShift systems, the namespace's service accounts must be
removed from the assigned SCCs. To find the namespace's service accounts that
are linked to SCCs, run the following code to generate the commands to review and
execute:SAS_NS=name-of-namespacefor oc_scc in $(oc get clusterrolebinding | grep "system:openshift:scc:" | awk '{ print $1 }'); do crb_name=$(oc get clusterrolebinding $oc_scc -o json | jq -r '.roleRef.name') sa_names=$(oc get clusterrolebinding $oc_scc -o json | jq -r '.subjects[] | select(.namespace == '\"$SAS_NS\"') | .name') for sa_name in $sa_names; do if [[ ! -z $sa_name ]]; then echo "oc -n $SAS_NS adm policy remove-scc-from-user ${crb_name##*:} -z $sa_name" fi donedone
Run each command that is produced as output from the code.
Remove All SAS Viya Platform Deployments in the
Cluster and the SAS Viya Platform Deployment
Operator
Remove the SAS Prepull and SAS
Workload Orchestrator ClusterRoles and
ClusterRoleBindings
Before you begin to uninstall your SAS Viya platform software, perform the following
tasks to account for some special cluster-wide permissions.
1If you enabled the Image Staging Node List Option in your deployment, follow
the instructions in the “Disable the Node List Option” section of the “SAS Image
|
the instructions in the “Disable the Node List Option” section of the “SAS Image
Staging Configuration Options” README located at $deploy/sas-bases/examples/sas-prepull/README.md
(for Markdown format) and $deploy/sas-bases/docs/sas_image_staging_configuration_options.htm
(for HTML
format).130 Chapter 6 / Uninstalling
|
2If your deployment includes Workload Management and Cluster Role was
enabled for it, follow the instructions in the “Disable the Cluster Role” section of
the ”Cluster Privileges for SAS Workload Orchestrator Service” README
located at $deploy/sas-bases/examples/sas-workload-orchestrator/configure/README.md
(for Markdown format) and $deploy/sas-bases/docs/configuration_settings_for_sas_workload_orchestrator_service.htm
(for
HTML format).
Remove the Deployments and the
Operator
To uninstall all the SAS Viya platform deployments in a cluster, repeat the
instructions in “Remove SAS Viya Platform from a Namespace” on page 129 for
each namespace in the cluster that contains a SAS Viya platform deployment. After
all the SAS Viya platform deployments have been removed, run the following
command from the $operator-deploy directory to remove the SAS Viya Platform
Deployment Operator:kustomize build . | kubectl -n name-of-deployment-operator-namespace delete -f -
Remove the Operator ClusterRoles
After all of the SAS Viya Platform Deployment Operator instances are removed from
the cluster, some shared cluster resources must be removed.kubectl delete clusterroles \ --selector "app.kubernetes.io/name=sas-deployment-operator"
Remove the Operator
CustomResourceDefinition
After all of the SASDeployment custom resources have been removed from the
|
Remove the Operator
CustomResourceDefinition
After all of the SASDeployment custom resources have been removed from the
cluster, the operator CustomResourceDefinition must be removed.kubectl delete crds \ --selector "app.kubernetes.io/name=sas-deployment-operator"
Additional Clean-up of Cluster Resources
Remove cluster-wide resources that have been left behind by the previous
commands.kubectl delete crd --selector "sas.com/admin=cluster-api"kubectl delete crd --selector "sas.com/admin=cluster-wide"kubectl delete clusterrole --selector "sas.com/admin=cluster-wide"
Uninstall with the SAS Viya Platform Deployment Operator 131
|
Remove SCCs
Note: If your deployment is not on Red Hat Openshift, skip this section.
If you have removed the last namespace running the SAS Viya platform in the
cluster, then the SAS-specific SCCs can be removed. The following can be run to
generate the commands to review and execute:
Note: Removing SCCs is a cluster-wide operation. Ensure that no namespaces
with SAS Viya platform software exist in the cluster before removing SCCs.
Otherwise, any remaining SAS Viya platform deployments in the cluster may cease
to work as expected.for sas_scc in $(oc get scc | grep sas | awk '{ print$1 }'); do echo "oc delete scc $sas_scc"; donefor sas_scc in $(oc get scc | grep pgo | awk '{ print$1 }'); do echo "oc delete scc $sas_scc"; done
Run each command that is produced as output from the code.
Validate the Removal
1After you have removed the last namespace with SAS Viya platform software
from the cluster, run the following commands:kubectl get crd --selector "sas.com/admin=cluster-wide"kubectl get crd --selector "sas.com/admin=cluster-api"kubectl get clusterrole --selector "sas.com/admin=cluster-wide"
If you removed the cluster-wide CRDs, each command should return zero
results. If you do get results, repeat the steps in “Additional Clean-up of Cluster
Resources” on page 131.
2If you are removing a deployment from OpenShift, run the following commands:kubectl get scc | grep saskubectl get scc | grep pgokubectl get clusterrole | grep "system:openshift:scc:sas-"
|
If you removed the cluster-wide SCC settings, each command should return zero
results. If you do get results, repeat the steps in “Remove SCCs” on page 132.132 Chapter 6 / Uninstalling
|
Remove the SAS Viya Platform
Deployment Operator Only
To remove SAS Viya Platform Deployment Operator from the cluster but leave the
instances of the SAS Viya platform running in their namespaces, run the following
commands:kubectl delete serviceaccounts,roles,rolebindings,deployments,secrets \ -n name-of-deployment-operator-namespace \ --selector "app.kubernetes.io/name=sas-deployment-operator"kubectl delete clusterrolebindings \ $(kubectl get clusterrolebindings \ --selector "app.kubernetes.io/name=sas-deployment-operator" \ -o jsonpath='{.items[?(@.subjects[0].namespace == "name-of-deployment-operator-namespace")].metadata.name}')
After the commands are completed, see “Remove the Operator ClusterRoles” on
page 131 for information to remove some remaining resources.
Uninstall Deployments That Do Not Use
the Deployment Operator
The instructions in this section should be used to uninstall a deployment that was
created using either Kubernetes commands (referred to as a "manual deployment"
in earlier versions of this document) or with the sas-orchestration command. The
steps assume that the SAS Viya platform is running and functional. If the software is
not running, start it. If the software is not functioning, contact SAS Technical
Support.
Remove the SAS Prepull and SAS Workload
Orchestrator ClusterRoles and ClusterRoleBindings
Before you begin to uninstall your SAS Viya platform software, whether from a
|
Orchestrator ClusterRoles and ClusterRoleBindings
Before you begin to uninstall your SAS Viya platform software, whether from a
single namespace or from an entire cluster, perform the following tasks to account
for some special cluster-wide permissions.
1If you enabled the Image Staging Node List Option in your deployment, follow
the instructions in the “Disable the Node List Option” section of the “SAS Image
Staging Configuration Options” README located at $deploy/sas-bases/examples/sas-prepull/README.md
(for Markdown format) and $deploy/sas-
Uninstall Deployments That Do Not Use the Deployment Operator 133
|
bases/docs/sas_image_staging_configuration_options.htm
(for HTML
format).
2If your deployment includes Workload Management and Cluster Role was
enabled for it, follow the instructions in the “Disable the Cluster Role” section of
the ”Cluster Privileges for SAS Workload Orchestrator Service” README
located at $deploy/sas-bases/examples/sas-workload-orchestrator/configure/README.md
(for Markdown format) and $deploy/sas-bases/docs/configuration_settings_for_sas_workload_orchestrator_service.htm
(for
HTML format).
Pause the SingleStore Operator
If your deployment includes SAS Viya with SingleStore, you must pause the
SingleStore operator before uninstalling the SAS Viya platform software.
1Pause the SingleStore cluster:kubectl -n name-of-namespace patch MemsqlCluster \ sas-singlestore-cluster --type=json -p \ '[{"op": "add", "path": "/spec/clusterState","value": "pause"}]
2When you pause the cluster, each leaf shuts down after snapshots are taken of
all the databases. The snapshot process can lengthen the shutdown time,
especially if a large amount of new data requires a snapshot. To check the status
of the snapshot process, view the log of the SingleStore pods. Here are some
examples:kubectl logs node-sas-singlestore-cluster-master-0 node \-n name-of-namespacekubectl logs node-sas-singlestore-cluster-leaf-ag1-0 node \-n name-of-namespace
During the snapshots process, the snapshot status is reported as a percent (%)
|
During the snapshots process, the snapshot status is reported as a percent (%)
complete. As shutdown completes, each leaf shows Completed Detaching
in
the log.
3When all of the pods terminate, proceed with uninstalling the SAS Viya platform.
Remove Service Account Links and SCCs
Note: If your deployment is not on Red Hat OpenShift, skip this section.
For Red Hat OpenShift systems, the namespace's service accounts must be
removed from the assigned SCCs.
1To find the namespace's service accounts that are linked to SCCs, run the
following to generate the commands to review and run:SAS_NS=name-of-namespace
134 Chapter 6 / Uninstalling
|
for oc_scc in $(oc get clusterrolebinding | grep "system:openshift:scc:" | awk '{ print $1 }'); do crb_name=$(oc get clusterrolebinding $oc_scc -o json | jq -r '.roleRef.name') sa_names=$(oc get clusterrolebinding $oc_scc -o json | jq -r '.subjects[] | select(.namespace == '\"$SAS_NS\"') | .name') for sa_name in $sa_names; do if [[ ! -z $sa_name ]]; then echo "oc -n $SAS_NS adm policy remove-scc-from-user ${crb_name##*:} -z $sa_name" fi donedone
Run each command that is produced as output from the code.
2If you have removed the last namespace running the SAS Viya platform in the
cluster, then the SAS-specific SCCs can be removed. The following can be run
to generate the commands to review and run:
Note: Removing SCCs is a cluster-wide operation. Ensure that no namespaces
with SAS Viya platform software exist in the cluster before removing SCCs.
Otherwise, any remaining SAS Viya platform deployments in the cluster may
cease to work as expected.for sas_scc in $(oc get scc | grep sas | awk '{ print$1 }'); do echo "oc delete scc $sas_scc"; donefor sas_scc in $(oc get scc | grep pgo | awk '{ print$1 }'); do echo "oc delete scc $sas_scc"; done
Run each command that is produced as output from the code.
3Run the following commands:kubectl get scc | grep saskubectl get scc | grep pgokubectl get clusterrole | grep "system:openshift:scc:sas-"
If you removed the cluster-wide SCC settings, each command should return zero
|
If you removed the cluster-wide SCC settings, each command should return zero
results. If you do get results, repeat the steps above to remove the SCCs or
ClusterRoles.
Remove Internal PostgreSQL Resources
Note: If your deployment does not include an internal instance of PostgreSQL, skip
this topic.
Remove the resources for an internal instance of PostgreSQL from the namespace:kubectl -n name-of-namespace delete postgresclusters --selector="sas.com/deployment=sas-viya"
Uninstall Deployments That Do Not Use the Deployment Operator 135
|
Note: If your Crunchy Data PVs do not have a reclaimPolicy of "retain", performing
this command will delete their associated PVCs, resulting in the data they contain
being lost. Therefore, ensure that you have backed up that data or are otherwise
unconcerned about losing it.
If your Crunchy Data PVs do have a reclaimPolicy of "retain", their PVCs will not be
deleted when the cluster is deleted. If you no longer need them, you can delete
them manually by using the kubectl get pvc
command to get the list of PVCs and
the kubectl delete
command to delete them.
Remove the SAS Viya Platform
Because your SAS Viya platform software is required to be deployed in a dedicated
namespace, uninstall your deployment by deleting that namespace. As an
administrator with cluster permissions, run the following command:kubectl delete namespace name-of-namespace
When the namespace is deleted, your SAS Viya platform deployment has been
uninstalled.
Note: If you have deployed the SAS Viya platform on multiple namespaces, repeat
the command on each namespace to remove the SAS Viya platform from the
cluster.
Remove CRDs
If you are removing the last namespace with a SAS Viya platform deployment from
the cluster, the SAS custom resource definitions (CRDs) must be removed as well.
Note: Removing CRDs is a cluster-wide operation. Ensure that no namespaces
with SAS Viya platform software exist in the cluster before removing CRDs.
|
with SAS Viya platform software exist in the cluster before removing CRDs.
Otherwise, any remaining SAS Viya platform deployments in the cluster may cease
to work as expected.
1As a user with rights to list CRDs, run the following two commands to determine
if any SAS CRDs still exist:kubectl get crd --selector "sas.com/admin=cluster-wide"kubectl get crd --selector "sas.com/admin=cluster-api"
2If either of those commands return objects, run the following two commands as a
user with rights to remove CRDs:kubectl delete crd --selector "sas.com/admin=cluster-wide"kubectl delete crd --selector "sas.com/admin=cluster-api"
3Run the following command as a user with rights to remove clusterroles:136 Chapter 6 / Uninstalling
|
kubectl delete clusterrole --selector "sas.com/admin=cluster-wide"
4Run the following commands:kubectl get crd --selector "sas.com/admin=cluster-wide"kubectl get crd --selector "sas.com/admin=cluster-api"kubectl get clusterrole --selector "sas.com/admin=cluster-wide"
If you removed the cluster-wide CRDs, each command should return zero
results. If you do get results, repeat the steps above to remove the CRDs.Uninstall Deployments That Do Not Use the Deployment Operator 137
|
138 Chapter 6 / Uninstalling
|
Appendix 1
(Optional) Using a Mirror Registry
Create a Mirror Registry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
About SAS Mirror Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
System Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Download Required Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Create a Basic Mirror Registry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Required Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Optional Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Using a Configuration File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Deploy the Software from the Registry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Create and Populate a Mirror Registry in Microsoft Azure . . . . . . . . . . . . . . . . . . . . . 147
Additional Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
|
Additional Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Configure Authentication for Azure Container Registry . . . . . . . . . . . . . . . . . . . . . . . . . 148
Create and Populate a Mirror Registry in Amazon ECR . . . . . . . . . . . . . . . . . . . . . . . . 149
Additional Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Using SAS Mirror Manager on AWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Create and Populate a Mirror Registry in Google Cloud Platform . . . . . . . . . . . . . . 150
Additional Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Create the Registry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Grant Permissions to the SAS Mirror Manager User . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Using SAS Mirror Manager on GCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Create and Populate a Mirror Registry in Red Hat OpenShift . . . . . . . . . . . . . . . . . . . 152
Create and Populate a Mirror Registry in JFrog Artifactory . . . . . . . . . . . . . . . . . . . . 153
|
Create and Populate a Mirror Registry in JFrog Artifactory . . . . . . . . . . . . . . . . . . . . 153
Create a Mirror Registry at a Dark Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Create a Mirror Registry
SAS Mirror Manager is a command-line utility that pulls SAS container images from
the SAS Container Registry and pushes the images to your container registry, which
can be either on-premises or in the cloud.139
|
A mirror registry is optional, but it supports specific use cases, such as sites with
limited internet access or those with a requirement to perform pre-deployment
security scanning of SAS packages. Setting up the mirror requires access to the
internet. However, after mirroring has been completed, the deployment can proceed
at sites that lack internet access.
About SAS Mirror Manager
In addition to enabling deployment from a pre-populated mirror registry, SAS Mirror
Manager supports other use cases. For example, you can use it to get a detailed list
of the available SAS Viya platform stable and cadence releases in the SAS
Container Registry. SAS Mirror Manager can help you determine whether
differences exist between a SAS Viya platform repository and the current contents
of the SAS Container Registry.
Using the SAS Mirror Manager list functionality, you can run checks before creating
a mirror registry for your organization. Here are some examples of those checks:
To display a list of all the available software cadences and versions for a SAS
software order:mirrormgr list remote cadences --deployment-data SASViyaV4_order-number_certs.zip
To display a list of all the available software versions and releases:mirrormgr list remote cadence releases --deployment-data SASViyaV4_order-number_certs.zip
To determine the required disk space for the available images:mirrormgr list remote repos size --latest --deployment-data SASViyaV4_order-number_certs.zip
|
System Requirements
The number of containers and the size of the images that you obtain from SAS
depend on the products in your software order. SAS Mirror Manager downloads
each image locally and then uploads it to the destination registry. The images that
are downloaded require approximately 30–50 GB of disk space on the machine on
which SAS Mirror Manager is running. You can use the --remove-after-upload
option with the mirrormgr mirror registry
command to remove container images
from the local cache after they have been uploaded to the destination registry.
However, make sure that you have enough space to store all of the concurrent
downloads.
You can run SAS Mirror Manager on 64-bit Linux, Windows, or macOS. By default,
SAS Mirror Manager performs the number of concurrent downloads that
corresponds to the number of available CPUs. You can override this setting with the --workers
option and a value that defines the number of concurrent threads to be
used for downloads. For more information, see “Required Flags” on page 144.
If you plan to use the SAS Viya Platform Deployment Operator with SAS Mirror
Manager, perform the steps to download SAS Mirror Manager and create the mirror
before deploying the operator. When you deploy the software for the operator, 140 Appendix 1 / (Optional) Using a Mirror Registry
|
download the certificates again. For more information, see “Deploy the SAS Viya
Platform Deployment Operator” on page 10.
SAS Mirror Manager does not support mirroring of Harbor versions 1.9.x and 1.10.x
because of a Harbor limitation. Harbor versions earlier than 1.9.0 are supported,
and Harbor version 2.0 and later are supported.
Before you run SAS Mirror Manager, perform the following steps:
nConfigure any relevant firewalls to enable secure outbound connections over
port 443 to the following locations: ses.sas.download
, ses.sas.com
, crbwp.sas.com
, bwp2.ses.sas.download
, and cr.sas.com
.
nIf you use a proxy server to facilitate outbound connections over the internet,
additional configuration is required:
oMake sure that the proxy server is configured to allow access to the SAS
sites that were listed previously.
oMake sure that the proxy server is not performing packet inspection.
SAS Mirror Manager uses mutual TLS for authentication. Authentication is
performed by sending a client certificate to the SAS servers. The required
client certificate is downloaded along with your SAS Software Order
deployment assets.
In a mutual TLS authentication scheme, the identity of the client (the client
certificate) is used to authenticate the client to the destination system. Packet
inspection by a proxy server entails decryption of traffic and the inability to
propagate the client certificate forward. Therefore, a packet-inspecting proxy
server is not supported.
|
propagate the client certificate forward. Therefore, a packet-inspecting proxy
server is not supported.
The proxy server is supported for use with SAS Mirror Manager as long as it
is not performing packet inspection and is configured for pass-through.
If you are unable to meet these requirements for firewalls and proxy servers,
follow the instructions in “Create a Mirror Registry at a Dark Site” on page 154.
nVerify network bandwidth.
SAS Mirror Manager requires a high-speed network environment. A lack of
sufficient bandwidth increases the amount of time that is required to create the
mirror registry and can lead to time-out errors when communicating with the
destination registry.
nDetermine the authentication mechanism for the target registry.
If you are using a cloud registry provider, use the corresponding CLI to access
the user name and password for your cloud registry. The user account that is
specified in the authentication must have permission to push images to the cloud
registry.
See the following documentation for more information about target registries that
require additional configuration:
nAzure Container Registry: ACR documentation
nAmazon Elastic Container Registry (Amazon ECR): Amazon ECR
documentation
nGoogle Cloud Platform Container Registry: GCR documentation
nRed Hat OpenShift Container Registry: OpenShift Container Platform registry
documentation
|
nRed Hat OpenShift Container Registry: OpenShift Container Platform registry
documentation
nJFrog Artifactory Registry: Artifactory Package Management documentationCreate a Mirror Registry 141
|
In a multi-tenant deployment, all tenant onboarding and offboarding YAML files must
include the transformer file site-config/mirror.yaml
. For more information, see
the README file at $deploy/sas-bases/examples/sas-tenant-job/README.md
(for Markdown format) or at $deploy/sas-bases/docs/onboard_or_offboard_tenants_using_the_tenant_job.htm
(for HTML format).
Download Required Files
To create a mirror registry with SAS Mirror Manager, start by obtaining the software.
When you order SAS software, SAS sends a Software Order Email (SOE) to your
business or organization. Your SOE includes information about the software order,
including a link to my.sas.com.
1Create a directory on your kubectl machine or on a machine that can be reached
by your kubectl machine:mkdir directory-name
SAS recommends that you name the directory deploy
, but you should use a
name that is meaningful to you. The directory is referred to as $deploy
in this
guide.
2Click the Get Started link that is provided in your Software Order Email (SOE).
The my.sas.com page opens in a web browser.
3Click My Orders near the top of the page.
The full list of your SAS software orders displays.
4Click the Deployment Tools button, which is located above the list of orders.
A popup window displays. It provides access to the SAS Viya Order API, SAS
Container Manager, and SAS Mirror Manager.
5Below the SAS Mirror Manager section, click Learn More in order to access the
SAS Mirror Manager page.
|
Container Manager, and SAS Mirror Manager.
5Below the SAS Mirror Manager section, click Learn More in order to access the
SAS Mirror Manager page.
The page opens in a separate browser tab.
6On the SAS Mirror Manager for SAS Viya page, click the download link in order
to download the appropriate SAS Mirror Manager package for the machine
where you want to create your mirror registry.
Note: If you have a version of SAS Mirror Manager that you used with an earlier
version of SAS Viya (such as SAS Viya 3.5), it is not compatible with the current
architecture of the SAS Container Registry.
7Return to the MySAS - Order List page in its separate browser tab.
8Click to select the order that you want to deploy.
9In the Viya Orders > Order Order ID pane that opens, examine the order
information. The description indicates the release cadence, the version of SAS
Viya platform that you are about to deploy, and whether it has been downloaded
previously. 142 Appendix 1 / (Optional) Using a Mirror Registry
|
10Click the Downloads tab near the top of the page.
The line above the Asset Type table indicates the release cadence and the
version of SAS Viya platform software that you will deploy.
11(Optional) If you want to deploy a different version, select the cadence and
version from the menus. Or scroll down to see previous versions for this order.
12Review the deployment documentation by clicking the How to Deploy link.
13If you are performing a manual deployment, skip this series of steps:
aCreate a new directory parallel to the $deploy
directory in which to store your
license and *-certs.zip files.mkdir directory-name
You should use a name that is meaningful to you. The directory is referred to
as $license
in this guide. Replace $license
with the directory name that
you prefer.
bSelect Deployment Assets from the Asset Type list, and then click the
Download button to download the files that are required to deploy your
software.
After unzipping the downloaded file, place the *-certs.zip file and the license
file, identified by the .jwt extension, in the $license
directory.
Note: If you plan to have multiple SAS Viya platform deployments in your
cluster, you should organize the $license
directory as you see fit. Whatever
strategy you use to organize, ensure that you can easily differentiate the
license and certs.zip files by order.
Save the TGZ file (the deployment assets) to the directory that you created in
step 1.
|
license and certs.zip files by order.
Save the TGZ file (the deployment assets) to the directory that you created in
step 1.
14Extract the files from the TGZ file in the same directory:tar xvfz file-name.tgz
The result is a directory structure that looks like this:$deploy/└── sas-bases/ ├── base/ ├── components/ ├── docs/ ├── examples/ └── overlays/
For a description of each subdirectory, see “$deploy/sas-bases Directory” on
page 27.
15Uncompress the SAS Mirror Manager file, mirrormgr- operating-system .tgz, in
your $deploy
directory.
16(Optional) SAS recommends saving the SOE in the same directory.Create a Mirror Registry 143
|
Create a Basic Mirror Registry
The following command creates and populates a mirror registry:mirrormgr mirror registry \ --destination myregistry.mydomain.com \ --username myregistryuser \ --password myregistrypassword \ --deployment-data ~path-to-certs-zip-file
The mirror registry
command performs the following tasks:
nPulls images from the SAS Container Registry.
nPushes the images to the specified private registry.
The optional --deployment-assets
flag is recommended. Without it, SAS Mirror
Manager picks up the latest release of all supported cadence versions, and
additional disk space is required to accommodate the software that is downloaded.
For more information, see “Optional Flags” on page 145.
If the target registry for the images requires an imagePullSecret for access to those
images, you will need to create one. For more information, see “Use
ImagePullSecrets to Access the Mirror Registry” on page 39.
Additional steps are required in order to create a mirror in each of the private cloud
environments that the SAS Viya platform supports. Follow one of the documentation
links below to set up the target registry:
n“Create and Populate a Mirror Registry in Microsoft Azure” on page 147
n“Create and Populate a Mirror Registry in Amazon ECR” on page 149
n“Create and Populate a Mirror Registry in Google Cloud Platform” on page 150
n“Create and Populate a Mirror Registry in Red Hat OpenShift” on page 152
|
n“Create and Populate a Mirror Registry in Red Hat OpenShift” on page 152
n“Create and Populate a Mirror Registry in JFrog Artifactory” on page 153
Required Flags
Here is the basic command syntax, which includes required flags:mirrormgr mirror registry required-flags
--deployment-data path-to-certs-zip-file
Specifies the full path to the certs.zip file that was downloaded to the$deploy
directory. You downloaded the ZIP file from my.sas.com.
--destination registry-location
Specifies the host name and repository path of the registry to which images are
replicated. Here is an example:registry.example.com/my-project
--password destination-registry-password
Specifies the password that grants you access to the destination registry.144 Appendix 1 / (Optional) Using a Mirror Registry
|
--username destination-registry-user-name
Specifies the user name that grants you access to the destination registry.
Optional Flags
--cacert path-to-CA-certificate
Specifies the CA certificate for communicating with the upstream repository
server. The default is $HOME/.config/mirrormgr/ca.pem
.
--cadence cadence-version
Specifies the cadence (or frequency) at which you deploy new SAS software. It
can be used with the -–release
flag. Here is an example of specifying a release
with the stable cadence, version 2020.1.1:--cadence stable-2020.1.1
Note: The --cadence
flag is not required. However, it enables you to mirror the
exact container image versions from which the accompanying deployment
orchestration assets were generated. By including the cadence and release, you
ensure that the mirror provides all the images that are required for a SAS Viya
platform deployment. If you have downloaded deployment assets from
my.sas.com, use the --deployment-assets
flag instead of the --cadence
flag.
--cert path-to-auth-certificate
Specifies the certificate for authenticating with the upstream repository server.
The default is $HOME/.config/mirrormgr/cert.pem
.
--config config-file-location
Enables you to automate the process of creating and populating a mirror registry.
The default value for the location is $HOME/.config/mirrormgr/mirrormgr.yaml
. For more information and an example, see “Using a
Configuration File” on page 147 .
-d , --debug
Enables debug-level logging.
|
. For more information and an example, see “Using a
Configuration File” on page 147 .
-d , --debug
Enables debug-level logging.
--deployment-assets
Uses data from the Deployment Assets archive that you downloaded with your
software order to determine the cadence name, version, and release values.
Replaces the --cadence
and --release
flags.
--latest
Mirrors only the latest versions of SAS content.
--log-file path-to-log-file
Specifies the log file location. The default is $HOME/.local/share/mirrormgr/mirrormgr.log
.
-p , --path path-to-repositories
Specifies the path where repositories are downloaded. The default is $HOME/sas_repos
.
Note: SAS recommends that you include this flag. Otherwise, the disk backing
the default home directory might fill up.Create a Mirror Registry 145
|
--platform platform-filter
Specifies a filter for identifying the repositories to mirror, by platform. Here is an
example of a platform filter:--platform x64-oci-linux-2
--push-only
Instructs SAS Mirror Manager to mirror the local copy to the specified registry
without connecting to the SAS Entitlement Service or to the SAS Container
Registry. This flag is only used along with the --destination
flag. For example
of the syntax, see “Create a Mirror Registry at a Dark Site” on page 154.
--release release-cadence-timestamp-ID
Specifies the software release that matches the deployment orchestration tools
that are used in your deployment assets file (in TGZ format). The release
includes the timestamp when the specified cadence version was last updated.
If the --release
flag is used, the –-cadence
flag must also be used. If the --release
flag is not used, SAS Mirror Manager mirrors the latest release of a
cadence version.
Note: You can derive the cadence, version, and release from the file name of
the deployment assets in the TGZ file that you download from my.sas.com. Here
is an example file name:SASViyaV4_09S468_stable-2020.1.1_20201016.1594946240211_deploymentAssets_2020-07-17T022620.tgz
In this example, stable-2020.1.1
corresponds to the cadence and version, and
the release includes a release year, month, and day plus an ID, such as 20201016.1594946240211
. However, if you have downloaded deployment
assets, you can use the -–deployment-assets
|
. However, if you have downloaded deployment
assets, you can use the -–deployment-assets
flag to pass in the cadence and
release automatically.
--remove-after-upload
Removes container images from the local cache after they have been uploaded
to a remote registry.
-r , --repo
Specifies a list of repositories to mirror. If not specified, all entitled repositories
are mirrored. To obtain a full list of repositories, run this command:mirrormgr list remote repos
--url URL
Specifies the base URL of upstream repositories. The default is https://
ses.sas.download/ses .
--workers number
Specifies the number of download threads to use. By default, the value is the
same as the number of cores on the local machine.
A complete list of flags used in the command is available in the Help for SAS Mirror
Manager. Access the Help using -h
or -–help
.146 Appendix 1 / (Optional) Using a Mirror Registry
|
Using a Configuration File
Passing in the location of a configuration file with the --config
flag lets you largely
automate the process of mirror creation and use a repeatable process. By default,
the location of the configuration file is assumed to be $HOME/.config/mirrormgr/mirrormgr.yaml
.
You can add virtually any of the available command flags to the configuration file by
removing the leading hyphens (--). The file must be in YAML format. Enclose
numeric string values in quotation marks (" ").
Here is an example of a valid configuration file:deployment-data: ./certs.zipcadence: "stable-2023.02"release: "20230217.1667851938227"path: ./sas_reposlog-file: ./mm.logdebug: true
Deploy the Software from the Registry
When you have verified that the SAS images to which you are entitled have been
added to the specified registry, you are ready to deploy the SAS Viya platform.
1Some updates to your kustomization.yaml file are required in order to configure
the deployment to use the mirror registry. For more information, see “Using a
Mirror Registry” on page 36.
2Your next steps depend on your deployment method:
nIf you are performing a deployment using the SAS Viya Platform Deployment
Operator, follow the steps in “Deploy the SAS Viya Platform Deployment
Operator” on page 10 in order to prepare your environment.
nIf you are performing a manual deployment, follow the steps in “Retrieve
Required Files” on page 18 to download additional files that are required for
|
Required Files” on page 18 to download additional files that are required for
the deployment. Then you can proceed to the Installation on page 31 steps.
Create and Populate a Mirror Registry in
Microsoft Azure
SAS Mirror Manager requires some additional configuration when the target
container registry is hosted on Microsoft Azure.Create and Populate a Mirror Registry in Microsoft Azure 147
|
Additional Requirements
Based on the average size of a SAS container deployment, a single deployment can
be performed from a Standard-sized Azure container registry. However, if you intend
to have a long-running registry that consumes SAS software updates over time, you
should select a Premium size. For more information, see the Microsoft Azure sizing
documentation.
Configure Authentication for Azure Container
Registry
Several methods of authenticating to the Azure Container Registry are available.
You can see a full list in the Azure container authentication docs. You can use the
Azure CLI to do the following:
1Log in to your Microsoft Azure account in order to authenticate your client: az login
2Get an authorization token:az acr login --name registry-name --expose-token
For registry-name , substitute the name of the destination container registry.
Note: The URL of any Azure container registry must be specified in all
lowercase letters.
The output that is returned includes the access token and the login server, as in
the following example:docker login loginServer -u 00000000-0000-0000-0000-000000000000 -p accessToken{ "accessToken": "token-contents", "loginServer": "mirrormgr.azurecr.io"}
3(Optional) The token might contain a large amount of data. It is a best practice to
save it as an environment variable:TOKEN=$(az acr login --name registry-name --expose-token --output tsv --query accessToken)
4Run SAS Mirror Manager with your credentials and append the token
|
4Run SAS Mirror Manager with your credentials and append the token
environment variable: mirrormgr mirror registry --deployment-data SASViyaV4_order-number_certs.zip --latest --destination registry-name.mydomain.com --username 00000000-0000-0000-0000-000000000000 --password $TOKEN
148 Appendix 1 / (Optional) Using a Mirror Registry
|
For registry-name.mydomain.com , specify the loginServer
that was returned
from the previous command as the --destination
. The URL must be specified
in all lowercase letters.
For access-token , specify the Azure token that was returned previously.
Note: The user name must be 00000000-0000-0000-0000-000000000000.
Return to “Required Flags” on page 144 for more information about the SAS Mirror
Manager commands. Consider whether you want to use any of the “Optional Flags”
on page 145. Then proceed to “Deploy the Software from the Registry” on page 147
for additional deployment instructions.
Create and Populate a Mirror Registry in
Amazon ECR
SAS Mirror Manager requires additional configuration when the target container
registry is hosted on AWS.
Additional Requirements
In order to support Amazon ECR as the mirror destination, make sure you have
fulfilled some additional requirements:
nThe AWS CLI must be installed.
nThe AWS CLI must be authenticated.
nThe user account must have permission to create new Amazon ECR
repositories. This permission is specified in the Identity and Access Management
(IAM) section of the AWS console.
Using SAS Mirror Manager on AWS
The image repositories must be created in Amazon ECR before they are pushed.
Take the following steps to use SAS Mirror Manager in an AWS environment:
1Create or configure an AWS private registry with Amazon ECR.
Your AWS account has an associated default private Amazon ECR registry. For
|
1Create or configure an AWS private registry with Amazon ECR.
Your AWS account has an associated default private Amazon ECR registry. For
more information, see Amazon ECR private registries .
2Follow the steps in “Download Required Files” on page 142 to obtain the SAS
Mirror Manager software.
3Create any required repositories:Create and Populate a Mirror Registry in Amazon ECR 149
|
for repo in $(mirrormgr list target docker repos --deployment-data SASViyaV4_order-number_certs.zip --destination namespace) ; do aws ecr describe-repositories --repository-names $repo --region region || aws ecr create-repository --repository-name $repo --region regiondone
Note: The call to describe-repositories
runs in order to determine whether
the repositories exist. If they do not exist, an error message is returned to alert
you about this condition. This error may be safely ignored.
For namespace , supply the name of the target namespace. The name of the
repository and namespace should conform to the repository name requirements
that are described in the Amazon ECR User Guide .
For region , supply the name of the AWS region. Be sure to use the same region
in all commands.
4Create and populate the mirror based on the associated software order:mirrormgr mirror registry -p ./sas_repos --deployment-data SASViyaV4_order-number_certs.zip --latest --destination registry-URL/namespace --username 'AWS' --password $(aws ecr get-login-password --region region)
The same options that are described in “Optional Flags” on page 145 are also
applicable to Amazon ECR.
If you encounter errors that report too many requests
, you can reduce the
number of worker threads. By default, the number of threads that is used is the
same as the number of CPU cores on the machine. Use the --workers
option to
set the number of threads.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.