SAS Visual Investigator 10.2 on the Cloud: Deployment Guide

Transcription

1 SAS Visual Investigator 10.2 on the Cloud: Deployment Guide

2 2

3 3 Introduction Deployment Architecture How SAS Visual Investigator Is Deployed System Requirements for SAS Visual Investigator Cloud Platform Software Requirements Virtual Machine Operating System and Software Requirements Web Browser Requirements Pre-installation Tasks Enable Required Ports Complete Configuration Worksheet Installing SAS Visual Investigator on SAS Viya Prepare the Installation Tools Prepare the Installation Environment Install SAS Visual Investigator Validating the Deployment Overview Contact SAS Technical Support Appendix A: Cloud Foundry Deployment Information Cloud Foundry Deployment Data Using vsphere as the IaaS Cloud Foundry Deployment Data Using OpenStack as the IaaS Appendix B: Troubleshooting BOSH Cloud Foundry Run Time

4 4 Introduction Deployment Architecture This document describes the processes to deploy SAS Visual Investigator in a private cloud. Only the Cloud Foundry private cloud environment is currently supported for cloud deployments. Instead of requiring Base SAS, SAS Visual Investigator uses the Cloud Analytic Services (CAS) run-time environment. How SAS Visual Investigator Is Deployed The installation and deployment of SAS Visual Investigator on Cloud Foundry uses a self-extracting installer to deliver the software into a directory on a physical server or virtual machine (VM). The installer creates a top-level directory named sas, which contains two subdirectories: the bin directory and the image directory. The bin directory consists of a single bash script named start. The image directory is a systemd-nspawn namespace container that is invoked by the start bash script and contains functionality that is used in SAS Visual Investigator deployment-related activities. It can be shut down and restarted whenever these functions are needed. In addition, the entire container, with any changes that are made, can be repackaged, archived, or moved to another server or VM as needed. You must perform these tasks to deploy SAS Visual Investigator: Prepare the installation tools. Prepare the installation environment. Install SAS Visual Investigator.

5 5 System Requirements for SAS Visual Investigator Cloud Platform Software Requirements Cloud Foundry Requirements Cloud Foundry v230 through v246 on either OpenStack or vsphere o OpenStack Mitaka version or later o VMware vsphere ESXi 6.0 U2 BOSH CLI or later cf CLI version d or later Stemcell CentOS 7,x 3309 sas_java_buildpack 3.12 Other Required Software The following third-party software is included with your SAS software: HashiCorp Consul Included to enable service discovery and configuration. RabbitMQ Included to provide an open-source, standards-based platform for SAS components and applications to send and receive messages. Elasticsearch Included to provide search capability. Apache Geode Used for session caching. Supported Databases The following database is supported for use by SAS Visual Investigator: PostgreSQL Here are the supported data stores: PostgreSQL 9.4 Oracle 12c Virtual Machine Operating System and Software Requirements The VM instance that is used to install SAS Visual Investigator has the following requirements: Red Hat Enterprise Linux or CentOS 7.2 must be installed. 20 GB free space must be available. The installation user must have sudo privileges. systemd-nspawn must be installed.

6 6 Web Browser Requirements The desktop machine that is used to access the SAS Visual Investigator user interface requires one of the following web browsers: Google Chrome 48 and later versions Microsoft Internet Explorer 11 Note: Microsoft Edge is not supported. Browsers on tablets and other mobile devices are not supported. SAS Environment Manager and CAS Server Monitor are not supported in this release of SAS Visual Investigator.

7 7 Pre-installation Tasks Enable Required Ports The following ports are used by SAS Visual Investigator and should be available before you begin to deploy your software. The same ports should also be available for any firewalls that are configured in the operating system or the network. Process Required Port Notes HTTPD 80 (internal) 443 (external) default Erlang Port Mapper Daemon (epmd) port 4369 SAS Infrastructure Data Server For a single server deployment with no failover, ports must be opened. Additional standby nodes each get the next available port number sequentially up to CAS Server Starting Port 5577 Used by clients to make binary connections to CAS. CAS Communicator Port 5580 default SAS Messaging Broker AMQP client access port 5672 SAS Configuration Server 8500 SAS uses HashiCorp Consul as its configuration server. Object Spawner 8591 CAS Server Monitor 8777 Used by clients to make REST HTTP calls to CAS, as with the Python REST interface. Elasticsearch 9200 Default PgPool port 9432 default SAS Messaging Broker management web console port SAS/CONNECT Spawner SAS Cloud Analytic Services Server default SAS Messaging Broker clustering port 25672

8 8 Process Required Port Notes Apache Geode Complete Configuration Worksheet To prepare for deployment, you must collect information about deployment via the key value pairs in the CF_Deployment_Questionnaire.xlsx spreadsheet. See index.html. Then, during installation, you must manually enter this information in a configuration file. To complete and customize your deployment, specify the information by associating required and optional keys with the appropriate values. Complete either the vsphere worksheet or the OpenStack worksheet, according to your environment. Specify a value under each Value column that is not already pre-populated. Do not specify any values for passwords. They are entered on-site. Blank lines have been inserted to improve readability and to show a separation between sections of the configuration file. JSON header lines are included to provide information about the section.

9 9 Installing SAS Visual Investigator on SAS Viya Prepare the Installation Tools Overview Installing SAS Visual Investigator begins by downloading the binary installer file to a physical or virtual machine (VM) in the IaaS environment. The VM is used as the bastion or jump box for the installation and deployment. The installation consists of the following tasks: Download the binary installer. Extract the files. Start the container. Sign on to the container. Download the Binary Installer Download the installer ZIP file that is included with the customer Software Order (SOE). For example, the file might look like visualinvestigator_10_2_xxx_lax.zip. Once the download is complete, the file must be unzipped. It looks similar to visualinvestigator_10_2_xxx_lax.bin. After downloading the installer ZIP file visualinvestigator_10_2_xxx_lax.zip from the location in the SOE, copy the file to the home directory of the installer ID on the bastion box using the wget command. The wget command is used to access a local web server: wget 10_2 xxx lax.zip :53: _2 xxx lax.zip Connecting to :80... connected. HTTP request sent, awaiting response OK Length: (5.1G) [application/octet-stream] Saving to: 'visualinvestigator 10_2 xxx lax.zip' 100%[=================================================>] 5,508,425, MB/s in 1m 43s :54:47 (50.8 MB/s) - 'visualinvestigator 10_2 xxx lax.zip' saved [ / ] Prepare the ZIP File Unzip the file visualinvestigator_10_2_xxx_lax.zip. Then, delete the ZIP file, visualinvestigator 10_2 xxx lax.zip. Run the visualinvestigator 10_2 xxx lax.bin file as follows: bash visualinvestigator 10_2 xxx lax.bin Extract the File Once the file is on the bastion box, to extract the binary file, run the following command:

10 10 bash visualinvestigator 10_2 xxx lax.bin Note: The binary installer deploys to the installing user s home directory. Start the Container Run the start.sh script in the user s home directory to start the container:./sas/bin/start Here are the displayed results: Starting SAS Visual Investigator Deployment Container... Spawning container <instance>.<hostname>.unx.sas.com on /home/test/sas/image. Failed to create directory /home/test/sas/image//sys/fs/selinux: No such file or directory Failed to create directory /home/test/sas/image//sys/fs/selinux: No such file or directory Press ^] three times within 1s to kill container. systemd 219 running in system mode. (+PAM +AUDIT +SELINUX +IMA -APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ -LZ4 -SECCOMP +BLKID +ELFUTILS +KMOD +IDN) Detected virtualization systemd-nspawn. Detected architecture x Welcome to CentOS Linux 7 (Core)! Because the container attempts to alter the state of the host operating system, the following error message might be displayed. This message is expected and can be safely ignored: Failed to create directory /home/test/sas/image//sys/fs/selinux: No such file or directory Failed to create directory /home/test/sas/image//sys/fs/selinux: No such file or directory [FAILED] Failed to start LSB: Bring up/down networking. See 'systemctl status network.service' for details. Sign On to the Container Once the container is started, a login prompt is displayed: CentOS Linux 7 (Core) Kernel el7.x86_64 on an x86_64 sfr47066 login: The prompt shows a host name with a randomly generated prefix and then the word login. Enter your user ID and then enter your password to sign on to the container. Last login: Wed Jun 15 12:41:34 on console SAS 15:04:47!1 [~] After you sign on to the container, notice that the prompt now includes the string, SAS, followed by a timestamp. For convenience, the shell is set to the Z shell (zsh), which provides the standard shell for SAS Visual Investigator deployments. Note: Do not change shells. Any other shell might produce unexpected results. The window of your X terminal emulator displays the title Container. The title is a reminder that your work environment is within the container and that you are no longer working within the host file system. Overview of Help Basic help: sas help

11 Provides information about the Help system itself and explains how to use the Help system. Package level How-to Guides: sas help -p Provides detailed instructions on how to use the package. In this case, instructions are provided on deploying SAS Visual Investigator. Component level help. sas help -c <component> Example: sas help -c deploy Commands in this section are used in the deployment and monitoring of services and applications into a Cloud Foundry environment. The Help system then repeats the same information displayed when using the command sas help sas. To show product information, run the following command: sas show info Product Name : SAS Visual Investigator Version: : Build Date: : Note: Additional information is displayed after running sas conf use <configuration>. 11 Prepare the Installation Environment Overview Before you can deploy SAS Visual Investigator, follow these pre-deployment steps: Collect information about the environment. Set, edit, and save the configuration file. Authenticate the sas user to Cloud Foundry and BOSH. Upload BOSH stemcells. Upload buildpacks. Upload releases. Generate files from configuration data for deployment. The information in the Microsoft Excel spreadsheet CF_Deployment_Questionnaire.xlsx needs to be collected. See Then, during installation, you must manually enter this information in a configuration file. Once the information is collected, enter the following command to list the available SAS Visual Investigator configuration files. sas conf list Here are the results: The following configurations are available: template_openstack template_vsphere To use a configuration, use the command: sas conf use <configurationname>

12 12 Current configuration: unset Set the Configuration File 1 Select the template file (vsphere or OpenStack) that corresponds with the underlying IaaS for your Cloud Foundry installation and run the following command: sas conf use template_xxxxx where xxxxx is either vsphere or openstack. Here is an example of the output: Now using configuration: template_xxxxx 2 Save the template configuration file with a new configuration filename that is meaningful in your environment. The configuration file test is used in this example. sas conf save test This produces the following output: Saved configuration file: test 3 Set the new configuration file as the active configuration file for the framework to use. sas conf use test This produces the following output: Now using configuration: test Edit the Configuration File Once the template is ready to use, edit the configuration file and enter the information from the spreadsheet. See Then, during installation, you must manually enter this information in a configuration file. Note: SAS has configured the vim utility to check JSON syntax for this purpose. You cannot exit the utility if the file contains invalid JSON syntax. Here is an example of the configuration file before editing:

13 13 After you enter the information into the table from Appendix A or the spreadsheet, copy your entries into the configuration file. 1 Copy your entries into the configuration file and ensure that your entries do not contain Rich Text markup. It must contain only ASCII 7 data. Every entry in Appendix A or the spreadsheet corresponds to an entry in the configuration file. 2 Save the configuration file. Here is an example of the configuration file after editing:

14 14 Note: The location of this configuration file that is in use is a temporary location and can easily be overwritten by another configuration file. SAS recommends that you save the configuration file that has been edited. Save the Configuration File To save the changes that you have made to a permanent location, run the following command: sas conf save test This produces the following output: Saved configuration file: test This command does not change the active configuration file. Instead, it saves a copy of it to a saved configuration directory. You can see your configuration name now by running the command: sas conf list This produces the following output: The following configurations are available:

15 15 test template_openstack template_vsphere To use a configuration, use the command: sas conf use <configurationname> Current configuration: test Authenticate to Cloud Foundry and BOSH Now that you have entered the information about your Cloud Foundry installation and the servers to be created, you need to connect to Cloud Foundry. This step is required even if the host bastion box is signed on to Cloud Foundry and BOSH. This step is also required whenever you change your configuration to a new configuration that has different authentication details. To authenticate, run the command: sas cf auth This produces the following output: Authentication for user: sas Current target is (ocfdir) API endpoint: (API version: ) User: test Org: test Space: test If authentication is successful, then you know that the data that you entered for the Cloud Foundry and BOSH environments is correct. If authentication is not successful, refer to the data that you entered for the directory and run-time sections of the configuration file. Correct them for your installation and try again. CAUTION! Do not attempt to proceed beyond this point if you do not have a working connection to both BOSH and Cloud Foundry. Upload the Buildpacks Deploying SAS Visual Investigator applications uses Cloud Foundry buildpacks. To ensure that compatible buildpacks are being used, run the following command: sas deploy buildpacks This produces the following output: sas deploy buildpacks Deploying buildpack sas_java_buildpack... OK Deploying buildpack sas_java_buildpack... Done uploading OK This command uploads offline buildpacks into your Cloud Foundry environment. Upload the BOSH Stemcells Once connectivity is established to the BOSH environment, you must upload the required stemcells for the creation of virtual machines (VMs). This is done in the IaaS using the BOSH Cloud Provider Interface (CPI).

16 16 1 Check to see whether the required stemcells are already installed. Run the following command: bosh stemcells If the output contains the following information, the correct stemcell is already installed. You can skip the remainder of this section. Name="bosh-openstack-kvm-centos-7-go_agent" Version="3309" 2 If the bosh-openstack-kvm-centos-7-go_agent stemcell is not installed, run the following command: sas bosh upload stemcells Two stemcells are delivered: one each for OpenStack and vsphere. 3 To ensure that the correct stemcells are loaded, even if there are existing stemcells, run the following command: sas bosch upload stemcells If the correct stemcells are already installed, you receive a warning message stating that the correct version already exists. You can safely ignore this message and continue with the upload. Upload the BOSH Releases Now you need to upload the included BOSH releases to the BOSH blob storage so that they are available to BOSH when it is time to deploy the services. To upload the BOSH releases, run the following command: sas deploy releases The following release packages are delivered, one for each of the stateful services. Cloud Analytic Server (CAS) Elasticsearch PostgreSQL 9.4 Consul RabbitMQ PGPool Apache Geode Create the Files for Deployment Once the stemcells, buildpacks, and releases are in place, you must create files that are used in deploying SAS Visual Investigator services and apps. This step should be performed each time there is a change in the active configuration file. If these files have already been generated, then remove them from the container first by running the following commands: cd rm -rf * The configuration file provides the values to be substituted into the template files that are part of the framework. To create the file system and files, run the following command: sas deploy files

17 As the command runs, it copies the JAR files that are deployed into Cloud Foundry. Then it performs the token substitution on the various files that are required for deploying the solution. When file system creation is complete, token substitution is checked to determine whether there are any unresolved tokens. If tokenization is successful, the message All tokens successfully resolved is displayed. If tokenization is incomplete and an error message is shown, the configuration file is incomplete and needs to be edited again. Note: Be sure to remove the files that were generated before running the command sas deploy files again. 17 Preparing to Examine the Files Examination of the BOSH and Cloud Foundry manifest files is required in order to verify that the information in the files is correct. You must have knowledge about the Cloud Foundry environment and the vim utility. Note: Do not change the information in the manifest files manually because the data comes from the configuration file. Any changes should be made to the configuration file. The contents of the home directory should then be deleted, and the command to deploy files should be rerun to create a new set of manifest files. Examine the Files Related to BOSH The manifest files that deploy stateful services into the BOSH managed environment are located in the /home/sas/services directory in the container. To examine the manifest files: 1 Change to the /home/sas/services directory. 2 Run the following command to display the vim editor: vi */man*.yml. This displays the vim editor in the window, with the first file in the vim buffer. Use the vim :n command to edit the next file match. Use the :rew command to rewind to the beginning if you want to review all files again. Note: It is acceptable to change the configurable entries in the jobs section of the manifest.yml file, which is in the Postgres directory under the services directory. They are intended to be managed outside the framework When examining the manifest files, check the following items: director_uuid make sure this matches what was entered in the configuration file. Make sure that all entries in the networks section of the file are correct. Validate that the IP address is the correct one entered for each server. Look under jobs and then static_ips. Once you have examined the BOSH manifest files, examine the script files in the services directory. To examine the script files: 1 Change to the /home/sas/services directory. 2 Run the following command to display the vim editor: vi */*.sh. This displays the vim editor in the window. The most important script in this group is the post_deploy_consul.sh script. In this file, check for empty values ("") and determine whether an empty value is reasonable. Some expected empty values include the archive.storage.local.destination property and in the Folder loop at the bottom of the file. A property with an empty value might indicate that a JSON property in the configuration file was not filled in. If you discover any entries like this, return to the configuration file and make sure that all required entries are filled in. Then delete the contents of the home directory and re-create the files. Note: The following service_tag parameters are blank for all environments: elasticsearch/deploy.sh service_tags, postgres/deploy.sh service_tags, and rabbitmq/deploy.sh service_tags. This is standard and can be ignored.

18 18 Note: If you are using OpenStack, the static_ips addresses in the cas-worker.yml and manifest-data.yml files are blank. This is standard for OpenStack and can be ignored. Examine the Files Related to Cloud Foundry Manifest files that deploy stateless applications and microservices into the Cloud Foundry run-time environment are located in the /home/sas/apps directory in the container. To examine the manifest files: 1 Change to the /home/sas/apps directory. 2 Run the following command to display the vim editor: vi */man*.yml In these files, check for empty values ("" ) and determine whether an empty value is reasonable. This might indicate that a JSON property in the configuration file was not filled in. If you discover any entries like this, return to the configuration file and make sure that all required entries are filled in. Then delete the directories under the /home/sas directory and re-create the files. Using Logs Log data for applications and microservices is provided through the Cloud Foundry Loggregator system. Using a Cloud Foundry firehose and nozzle to collect log data and redirect it to a logging server is the standard means of providing access to application logs. The rsyslog log has been added to the services deployed in Cloud Foundry BOSH. This information can be routed to any rsyslog server over TCP. This method is compatible with any of the various log store and display applications. Install SAS Visual Investigator Overview Deployment of SAS Visual Investigator consists of the following: deploying services to BOSH deploying stateless applications and microservices Deploy Services to BOSH To deploy the stateful services to BOSH, run the following command: sas deploy services When deployments are complete, test the applications to ensure that the deployment is working as expected. For more information, see Validating the Deployment on page 21. CAUTION! Do not attempt to proceed to the next steps if you see any errors during the deployment of the stateful services. If you see errors, see Appendix B: Troubleshooting on page 56. Deploy Stateless Applications and Microservices To deploy the stateless applications and microservices, run the following command: sas deploy apps

19 This is similar to the command that deploys the stateful services. The sas deploy apps command runs the deploy.sh script in the /home/sas/apps directory. This script executes each of the deploy.sh scripts in the subdirectories under the apps directory. These scripts deploy application JAR files into the Cloud Foundry runtime environment. When the deployments are complete, run the following command to display the status of the full deployment: sas show status Note: The preceding command combines three different commands bosh deployments, cf apps, and cf routes to display information that describes the current environment. Also, use the following command to display information about the applications that are deployed. sas show info It also displays the URLs that enable you to connect to the application interfaces, and it displays information about the environment. All of the information displayed comes from the configuration file and is formatted for readability. Here is typical output: SAS Visual Investigator: Info Product Name : SAS Visual Investigator Version: : 10.1.nnn Build Date: : Day Mon YY HH:MM:ss EDT Configuration JSON Configuration Org Space Host URLs SAS Visual Investigator Consul RabbitMQ : test : test : test : test : : : IP Addresses CAS Controller : CAS Worker : Consul : Elastic Search : Master Nodes : Client Nodes : Data Nodes : PostgreSQL : RabbitMQ : Another way to quickly display the status of a stateful service or application is to use the following command: sas show details Here is typical output: Service Status pass cas-controller (Success) pass cas-worker@ (Success)

20 20 pass (Success) pass consul_container (Agent alive and reachable) pass elasticsearch (TCP connect :9200: Success. ElasticSearch cluster green.) pass postgres (Success) pass rabbitmq (Success) App Status pass audit (UP) pass authorization (UP) pass casmanagement (UP) pass datahub (UP) pass entityresolution (UP) pass feature (UP) pass files (UP) pass identities (UP) pass networkanalytics (UP) pass SASLogon (UP) pass SASVisualInvestigator (UP) pass svi-ai (UP) warn svi-alert (DOWN) pass svi-core (UP) pass svi-sand (UP) pass svi-transport (UP)

21 21 Validating the Deployment Overview This section provides instructions for validating the stateful services that are delivered with SAS Visual Investigator. Elasticsearch To validate the installation of the Elasticsearch cluster, run the following command: curl -XGET ' Elasticsearch-IP-address:9200/_cluster/health?pretty=true' Here is typical output: { } "cluster_name" : "testcluster", "status" : "green", "timed_out" : false, "number_of_nodes" : 2, "number_of_data_nodes" : 3, "active_primary_shards" : 5, "active_shards" : 10, "relocating_shards" : 0, "initializing_shards" : 0, "unassigned_shards" : 0, "delayed_unassigned_shards": 0, "number_of_pending_tasks" : 0, "number_of_in_flight_fetch": 0, "task_max_waiting_in_queue_millis": 0, "active_shards_percent_as_number": 100 The green status value indicates that the cluster s overall operation is good. For additional information about Elasticsearch cluster health, see For additional Elasticsearch information, see Consul To validate the installation of Consul: 1 Open a web browser and go to the URL for Consul. 2 To determine the correct URL for Consul, run the following command: sas show info SAS Visual Investigator: Info Product Name : SAS Visual Investigator Version: : Build Date: : Wed Jun 22 17:48:41 EDT 2016

22 22 Configuration JSON Configuration Org Space Host URLs SAS Visual Investigator Consul RabbitMQ : test : test : test : test : : : IP Addresses CAS Controller : CAS Worker : Consul : Elastic Search : Master Nodes : Client Nodes : Data Nodes : PostgreSQL : RabbitMQ : From the preceding output, select the URL for Consul. In this example, the URL is :8500. RabbitMQ To validate the installation of RabbitMQ: 1 Open a web browser and go to the URL for RabbitMQ. 2 To determine the correct URL for RabbitMQ, run the sas show info command. From the previous example, the correct URL would be PostgreSQL The simplest way to validate the installation of the PostgreSQL server is to connect to a remote database. To determine the IP address of the PostgreSQL server, run the sas show info command. From the previous example, the IP address would be Then run the following command to validate the PostgreSQL server: psql -h U dbmsowner postgres Password for user dbmsowner: psql (9.2.15, server 9.4.4) WARNING: psql version 9.2, server version 9.4. Some psql features might not work. Type "help" for help. postgres=# To exit, specify \q and then press the Enter key.

23 23 Contact SAS Technical Support Technical support is available to all customers who license SAS software. However, we encourage you to engage your designated on-site SAS support personnel as your first support contact. If your on-site SAS support personnel cannot resolve your issue, have them contact SAS Technical Support to report your problem. Before you call, explore the SAS Support website at support.sas.com/techsup/. This site offers access to the SAS Knowledge Base, as well as SAS communities, Technical Support contact options, and other support materials that might answer your questions. When you contact SAS Technical Support, you are required to provide information, such as your SAS site number, company name, address, and phone number, that identifies you as a licensed SAS software customer.

24 24 Appendix A: Cloud Foundry Deployment Information Cloud Foundry Deployment Data Using vsphere as the IaaS SYNTAX CONVENTIONS #========================================================================== Syntax Description " " A quoted string of the indicated type. [ ] An array of comma-separated values of the indicated type. <name> *<size>* All occurrences of <name> can be replaced with the same replacement value. All occurrences of *<size>* can be replaced with a similar value or a larger value. The exact size depends on how the resource is used. For a size value in the following configuration table, ensure that the value for your instance_type definition is not less than the values that is used for CPU, disk, or RAM as follows: Size Property *<small>* *<medium>* *<large>* *<xlarge>* *<2xlarge>* Specify the Default Size or a Large Size, as Appropriate 2 vcpus, 20 GB disk, 2 GB RAM 4 vcpus, 40 GB disk, 4 GB RAM 4 vcpus, 40 GB disk, 8 GB RAM 8 vcpus, 40 GB disk, 8 GB RAM 8 vcpus, 40 GB disk, 8 GB RAM # A 'property' is a simple key-value pair. # A 'section' is either an array or a dictionary that can have many values. # # All properties and sections below must be populated with data unless marked with one of the following indicators: ## Optional - This property or section can remain with no value or can be entirely removed. ## ReadOnly - Do not change any property or section that has this mark. ## ReadOnlyKey - Do not change the property key, but you can change the value. The Value column does not list a default value in some instances. This is intentional so the user can provide customized values. BOSH Director

25 25 directors <directorname> username username password certificate_path privateip uuid vm_network_name bosh_domain_name dnsserver[ ] gateway cidr_network reserved_network static_network The BOSH Director section of the JSON configuration file provides information for accessing the BOSH directors. The director Name that is displayed by the bosh status command and that is listed as Name. The BOSH user name that is used for BOSH director authentication. The BOSH user name that is used for BOSH director authentication. The password for the user that connects to the BOSH director. The certificate for the user that connects to the BOSH director. The IP address of the BOSH Director that is displayed by the bosh status command and that is under the URL. The UUID of the BOSH director that is displayed by the bosh status command and that is listed as uuid. The vsphere network name under which the BOSH instances are created. The domain name of the BOSH director that is displayed by the bosh status command and that is listed as DNS. The array of quoted IP addresses that identify the DNS locations for the BOSH director. Example: [" ", " "]. The IP address of the Gateway server of the BOSH Director. The CIDR range of the network. Example: " /24". The reserved IP address range that is not used for creating the BOSH instances. The format of each array entry is a quoted string that consists of a hyphenseparated IP address range. Example: [" ", " "]. The static IP address range that is used for creating the BOSH instances. The format of each array entry is a quoted string that consists of a hyphen-separated IP address range. Example: [" ", " "]. Cloud Foundry Run-Time Environment

26 26 runtimes <runtimename> url username password The Cloud Foundry Run-Time Environment section of the JSON configuration file provides information for accessing the Cloud Foundry run-time environment. The name of the run-time environment that is displayed by the bosh deployments command and that is listed as Name. The URL of the Cloud Foundry run-time environment that is displayed by the cf api command. The Cloud Foundry run-time environment admin user name that is used for authentication. This user must have admin privileges in order to create domains and routes. The password for the user connection to the Cloud Foundry run-time environment. options --skip-ssl-validation The command-line options for the Cloud Foundry runtime environment connection. Solution Deployment deployments sas password The Solution Deployment section of the JSON file contains solution deployment details. This section provides information about the user ID that is used for solution deployment. The password for the user of the solution deployment. Do not provide a password in this file. It is entered on-site. iaas vsphere The Infrastructure as a Service (IaaS) name for vsphere. runtime <runtimename> The name of the Cloud Foundry run-time environment that is defined in the Cloud Foundry Run-Time Environment section. director <directorname> The name of the Cloud Foundry BOSH director that is defined in the directors section. The following three properties specify the top-level domain (TLD) that is a Cloud Foundry Shared or Private domain. Note: All three of these properties must resolve to the Cloud Foundry HAProxy. cf_nontls_domain front_tls_domain This is the internal non-tls domain that resolves directly to the Cloud Foundry HAProxy. If TLS front-end reverse proxying or load balancing is added by the customer, then it is necessary to specify the front-end TLS domain name. If used, this top-level domain can indirectly resolve to the Cloud Foundry HAProxy.

27 27 front_nontls_domain label org space If non-tls front-end reverse proxying or load balancing is added by the customer, then it is necessary to specify the front-end non-tls domain name. If used, this top-level domain can indirectly resolve to the Cloud Foundry HAProxy. This property identifies a unique deployment. The use of label in this context refers to a word or set of characters that is placed before the top level domain name (non-tls Domain name). This name immediately precedes the toplevel domain property following in Logging Properties > Domain cell. This name can be used to identify the deployment purpose, the function, or the intended audience of the deployment. It can also be a name that has no obvious meaning. From a deployment perspective, this name is needed to uniquely construct URLs that access a specific deployment. The name of the Cloud Foundry organization that is the target of the deployment. The name of the Cloud Foundry space that is the target of the deployment. LDAP Server ldap type base host password port userdn The LDAP Server section of the JSON file provides information about the defined LDAP AD connection. The string must match either ActiveDirectory or OpenLDAP. This section provides base OU information for the LDAP AD (Active Directory) connection. The LDAP server host name. The credentials for the user ID of the solution installation that connects to the LDAP/AD server. The credentials are entered on-site. Do not enter passwords in this file, they will be entered on-site. <Password terminology>. The port that is defined to connect to the LDAP server. If a global port is used to connect to the LDAP server, then port 3268 is required. If a non-global port is used to connect to the LDAP server, then port 389 is required. The LDAP user DN value that is used for the general search account. The DN value can be in the form: "CN=, OU=, OU=, DC=, DC=, DC=". Note that each site is different. LDAP Group Information group The LDAP Group Information section of the JSON file specifies group information for the LDAP/AD connection.

28 28 basedn searchfilter The LDAP base DN in the form OU=Groups,DC=xxx,DC=yyy,DC=com. The optional LDAP search filter that is used for searches on the LDAP server. Example: samaccountname={0}. LDAP Connection user basedn searchfilter The LDAP Connection section of the JSON file specifies user information for the LDAP/AD connection. The LDAP base DN is in the form DC=xxx,DC=yyy. The optional LDAP search filter that is used for searches on the LDAP server. Example: samaccountname={0}. Internal SAS profile The Internal SAS section of the JSON file is internal to SAS. file ldap/ldap-search-and-bind.xml Use the provided information that is required internal to SAS microservice. Connection Properties property <property name> "config/application/ldap/user/ customfilter": "" The Connection Properties section of the JSON file specifies any special properties that are required for connection to the LDAP/AD server. Define the custom Consul properties for LDAP. At the present time, the only supported property is the customfilter property that is listed. This property can be entirely omitted if no custom Consul properties are needed for LDAP. If this property is used, then the key must match config/application/ldap/user/customfilter: " ". Example: "(&(!(objectclass=computer))(! (useraccountcontrol: :=2)))." Tenancy tenancy userrdn The Tenancy section is for information that is required for managing tenancy through the LDAP/AD server. Example "ou=people". Consul Scripts oauth adminsecret The Consul Scripts section of the JSON file is used by scripts that load data into Consul. The password is used to obtain an access token from SASLogon for client registration. Do not provide a password in this file. It is entered on-site. Logging Properties

29 29 log_server host The Logging Properties section of the JSON file defines logging properties that are used by services that send log data output to a remote rsyslog server via TCP. To enable logging support for apps, a nozzle must be created that logs output to a remote rsyslog server. A nozzle is a component that listens for specified events and metrics and streams this data to external services. Refer to the Cloud Foundry Loggregator documentation for details about nozzles. Cloud Foundry nozzle and the rsyslog server must exist prior to the deployment. This functionality is not included with this deployment code. The fully qualified domain name (FDQN) or the IP address of the rsyslog server in the format: <host>.<toplevel-domain>. Example: rsysloghost.mycompany.com port 5000 The TCP port number of the rsyslog server. Example: Change the value, as appropriate. domain This is the top-level domain name of the rsyslog server. This is only <top-level-domain>, not <host>.<top-leveldomain>. Example: mycompany.com. Fallback Logging log_dir /log This section of the JSON file is used to define logging. If the log_server section for logging is not used, a default logging mechanism is available for apps only. This property defines the base directory to which logs are written. The directory must exist if this type of logging is used. If this directory does not exist, then logging is not enabled. If the directory exists, then subdirectories in the form of <organization>/<space>/<deployment> are automatically created, and log files are placed in that directory. For supporting apps only, the default logging mechanism uses cf logs to obtain data. Note that the default logging mechanism does not apply to services. Also note that the pathname /log is an example. Change the pathname, as appropriate. You might use a mount point to an NFS location. Be sure to add the mount point to /etc/ fstab so that subsequent sessions perform automatic mounting. SAS recommends that you use the log_server approach since this provides the best results. Doing so allows the user to see the logs from their services as well as the applications. This approach is more CPU-intensive and does not produce logging data for all deployment components. Use the "log_dir" approach for simple logging with the restrictions mentioned above. log_timeout 15m If logging is performed using log_dir, then terminate logging after this time-out value. If the value is "0", then logging will not time out. If this value is omitted, then a default time-out is used. The value is an integer immediately followed by either 's' for seconds, 'm' for minutes, 'h' for hours, or 'd' for days. BOSH Services

30 30 service default The BOSH Services section of the JSON file contains information about Bosh services. The default values are examples only. They provide a starting point for all of the same items in a section. This allows the software to have a standard set of basic items that can be overridden by the user if required for all items of the same type. Consult with your BOSH administrator for values that are appropriate for your deployment. The default values are used across all BOSH services. A service that does not define the properties that are specified in the default section use these properties. schema v1 Use this setting to specify the BOSH manifest schema version allowed options are v1 and v2. Cloud Config cloud_config vm_type network_name The next two properties are specific to BOSH manifests using the v2 schema. Select the cloud-config name that specifies the VM size configuration that will be used in your BOSH v2 manifests. Example: "default" Select the cloud-config name that specifies the network properties that will be used in your BOSH v2 manifests. Example: "bosh_services_network" canaries 1 The number of test VMs that are created before the creation of actual VMs for use. BOSH attempts to upgrade a small number of VMs (usually 1) in a batch. This first batch is called the canary. If the upgrade is successful, the remaining servers are created or upgraded, as appropriate. max_errors 1 The maximum number of compilation errors to be tolerated before a deployment fails. The recommended value is 1. max_in_flight 4 The maximum number of non-canary instances to create or update in parallel. This is the default number of instances to be created. The number can be modified later, as necessary. nfs_mounts [ ] A comma-separated array of NFS mount descriptors. Each descriptor is in the form: <hostname>:<mountdirectory-from-host>:<mount-directory-in-container>. NOTE: A single mount point for the SAS license file is required. The final entry /local/directory in the nfs_mount value must be /opt/cas/license. This value is required in order for CAS to locate the SAS license file. Example 1: "my.host.com:/remote/directory:/opt/cas/ license,other.host.com:/dir:/opt/cas/license". Example2: "my.host.com:/remote/directory:/opt/cas/ license,other.host.com:/dir:/opt/cas/license"

31 31 watch_time The default watch_time value ranges from compilation This section relates to the compilation phase of job deployment on a VM that is being created. The sizes used in this section should be no smaller than *<2xlarge> (8 vcpus, 40GB disk, 16GB RAM)* settings. workers 2 The number of workers that are specified in the deployment configuration determines how many VMs can be created at once for compiling. reuse_vms TRUE BOSH creates a new compilation VM for each new package compilation and destroys the VM when compilation is complete. If the value is TRUE, then compilation VMs are reused when compiling packages. The value, TRUE, is recommended. CPU 2 The number of vcpus to be assigned to the compilation VM. As a default for the compilation VM, use a 2xlargesized VM. The number of vcpus in the instance_type definition might be smaller than the 8 vcpus used for the 2xlarge value. disk The size of the disk of the VM to be created for compilation. ram The amount of RAM to be assigned to the VM for compilation. Resource Pool resource_pools A collection of VMs that are created from the same stemcell, with the same configuration, in a deployment. CPU 2 The number of vcpus to be assigned to the Resource Pool. As a default for the VM to be instantiated, use a large-sized (4 vcpus, 40GB disk, 8GB RAM) VM. The number of vcpus in the instance_type definition might be smaller than the 4 vcpus used in the "large" value. disk The size of the disk of the VM to be created for the Resource Pool. ram 8192 The amount of RAM to be assigned to the VM for the Resource Pool. Jobs jobs The Jobs block defines how BOSH associates jobs with the VMs started by the IaaS. If the service requires more disk space than is shown below, be sure to create an override in the specific section to allocate more disk space.

32 32 persistent_disk The specification of the persistent disk size. BOSH creates a persistent disk of the specified size in megabytes and attaches the disk to each job instance VM. The default for SAS is 20GB. Change the value according to your available or required resources. High- Availability Support ha enabled When HA support is enabled, the following configuration sections are used for the deployment: Consul: "consul_ha" PostgreSQL: "postgres_ha" RabbitMQ: "rabbitmq_ha" Geode: "geode" Enable HA support. Enable or disable HA support. The default is "false", which indicates that HA is disabled. SAS in-memory Server cas_controller The SAS in-memory Server section of the JSON file specifies the license controller node information. uses_workers FALSE This property turns off the ability of the license server to use worker nodes. This is the default for this server. super-user ip This is the CAS user that can administer all CAS sessions. The IP address to be used for the controller node. port 5577 The port to be used for the controller node. The required port is service_name license This is the prefix that will be applied to the service name in Consul. Worker Nodes cas_worker ip [ ] This section of the JSON file specifies worker node information for the SAS Cloud Analytic Server (CAS). Array of comma-separated IP addresses. One for each node that will be deployed. If cas_controller.uses_workers is FALSE, then this section is ignored. Consul consul ip The Consul section of the JSON file specifies Consul information that is provided by SAS. The IP address to be used for Consul.

33 33 High-Availability (HA) Consul consul_ha virtual_ip ip [ ] cloud_config network_name Information for High-Availability (HA) Consul. The virtual IP address is the IP that will be exposed in the Consul instances. The IP addresses to use for the Consul instances. The following properties are used if using the cloudconfig (BOSH manifest schema v2). Select the cloud-config name that specifies the manual network properties. Elasticsearch Master elasticsearch_master ip [ ] The Elasticsearch Master section of the JSON file specifies information that is provided by SAS for Elasticsearch master nodes. The array of comma-separated IP addresses, one for each node to be deployed. port 9200 The port that is used by Elasticsearch. Do not change this value. resource_pools The override value for the default value for the resource_pools property. The size should be no smaller than the *<2xlarge> (8 vcpus, 40GB disk, 16GB RAM)* settings. ram The override for the memory used by this service. Do not use a smaller value. properties These are additional properties that are made available to Elasticsearch. Refer to the Elasticsearch documentation for details. threadpool_bulk_queue_size 1000 This property value is for the bulk request. This tells Elasticsearch the number of requests that can be queued for execution in the node when no thread is available to execute a bulk request. min_heapsize max_heapsize Set the minimum JVM heap size. Follow Elasticsearch recommendations for setting the minimum JVM heap size. Set the maximum JVM heap size. Follow Elasticsearch recommendations for setting the maximum JVM heap size. Elasticsearch Client