These Terraform modules allow you to create and provision infrastructure for PeopleSoft on Oracle Cloud Infrastructure using Terraform. Peoplesoft can be deployed on Oracle Cloud Infrastructure in single availability domain or multi availability domain architecture.The modules can be used to create infrastructure for Peoplesoft in single Availability Domain as well as multiple Availability Domains.
For information on Oracle Peoplesoft deployment architecture on Oracle Cloud Infrastructure, see
- Architecture for Deploying PeopleSoft in a Single Avalability domain
- Architecture for Deploying PeopleSoft in Multiple Availability Domains
- Download and install Terraform (v0.11.8 or later)
- Export OCI credentials using guidance at Export Credentials.
- The tenancy used for provisoning must have service limits increased to accomodate the build.
Terraform modules for PeopleSoft has the following folder structure:
.
├── datasources.tf
├── _docs
│ ├── peoplesoft_multiple_availability_domain_ha_topology.png
│ └── peoplesoft_single_availability_domain_ha_topology.png
├── env-vars
├── env-vars.ps1
├── LICENSE.md
├── main.tf
├── modules
│ ├── bastion
│ │ ├── bastion.outputs.tf
│ │ ├── bastion.tf
│ │ └── bastion.vars.tf
│ ├── compute
│ │ ├── compute.blockvolume.tf
│ │ ├── compute.datasources.tf
│ │ ├── compute.outputs.tf
│ │ ├── compute.remote-exec.tf
│ │ ├── compute.tf
│ │ └── compute.variables.tf
│ ├── dbsystem
│ │ ├── db.datasources.tf
│ │ ├── db.dbsystem.tf
│ │ └── db.variables.tf
│ ├── filesystem
│ │ ├── fss.data.tf
│ │ ├── fss.outputs.tf
│ │ ├── fss.tf
│ │ └── fss.vars.tf
│ ├── loadbalancer
│ │ ├── lb.tf
│ │ └── lb.vars.tf
│ └── network
│ ├── subnets
│ │ ├── subnets.output.tf
│ │ ├── subnets.tf
│ │ └── subnets.vars.tf
│ └── vcn
│ ├── vcn.data.tf
│ ├── vcn.outputs.tf
│ ├── vcn.tf
│ └── vcn.vars.tf
├── outputs.tf
├── provider.tf
├── README.md
├── routetables.tf
├── seclist.tf
├── terraform.tfvars
├── userdata
│ ├── bootstrap.tpl
│ └── init.sh
└── variables.tf
11 directories, 41 files
-
[root]:
- [env-vars]: This is an environment file to set terraform environment variables on UNIX systems.
- [env-vars.ps1]: This is an environment file to set terraform environment variables on Windows systems.
- [datasources.tf]: This is terraform data source file to fetch data for Oracle Cloud Infrastructure resources.
- [main.tf]: At root level, main.tf calls different modules to create Oracle Cloud Infrastructure resources.
- [outputs.tf]: This is the terraform outputs file.
- [provider.tf]: This is the terraform provider file that defines the provider (Oracle Cloud Infrastructure) and authentication information.
- [variables.tf]: This is the terraform variables file to declare variables.
- [routetables.tf]: This file creates route tables.
- [seclists.tf]: This file creates security lists.
- [terraform.tfvars]: This is an input file to pass values to declared variables.
-
[modules]: The modules directory contain all the modules required for creating Oracle Cloud Infrastructure resources.
- [bastion]: This module is used to create bastion hosts.
- [compute]: This module is used to create unix and windows compute instances.
- [dbsystem]: This module is used to create Oracle Cloud Infrastructure database system.
- [filesystem]: This module is used to create Oracle Cloud Infrastructure File Storage service filesystem.
- [loadbalancer]: This module is used to create Oracle Cloud Infrastructure load Balancing service.
- [network]: This module is used to create network resources like VCN (Virtual Cloud Network),subnets, internet gateway, service gateway, dynamic routing gateway and NAT (network Address Translation) gateway.
-
[vcn]: This sub module creates the VCN, internet gateway, service gateway, dynamic routing gateway and NAT gateway.
-
[subnets]: This sub module create subnets within a VCN.
-
[userdata]: This directory contains bootstrap and shell scripts.
- [bootstrap.tpl]: This file contains cloud-init configuration to bootstrap compute instances.
- [init.sh]: This is an optional sample bash script file to initialize compute instances.
-
The following inputs are required for this module:
| Argument | Description |
|---|---|
| AD | Availability Domain for deployment. This variable drives the architecture to be deployed. Setting AD = ["1"] deploys infrastructure in single availability domain (Availabilty domain 1 of the tenancy in this case) and AD = ["1","2"] deploys infrastructure in multiple ADs (Avilability domains 1 and 2 of the tenancy in this case). |
| vcn_cidr | CIDR block of the VCN (Virtual Cloud Network) to be created. |
| vcn_dns_label | DNS Label of the VCN (Virtual Cloud Network) to be created. |
| linux_os_version | Operating system version of Oracle Linux for compute instances. The terraform module for compute instances always pick up the latest image available for the chosen Oracle Linux version in the region. |
| timezone | Timezone of compute instances. |
| bastion_user | User name to log in to bastion host. |
| compute_boot_volume_size_in_gb | Size of boot volume in gb for compute instance. |
| compute_block_volume_size_in_gb | Size of block volume in gb for compute instance. |
| compute_instance_user | User name to log in to compute instance. |
| psft_env_prefix | Environment prefix to define names of Peoplesoft resources. |
| psft_app_instance_count | Number of Peoplesoft application server instances to be created. |
| psft_app_instance_shape | Shape of Peoplesoft application instance. For more information on available shapes, see VM Shapes |
| psft_app_instance_listen_port_range | Port on which Peoplesoft application instance will listen and receive requests from Load Balancer. |
| psft_ps_instance_count | Number of Peoplesoft process scheduler instances to be created. |
| psft_ps_instance_shape | Shape of Peoplesoft process scheduler instance. For more information on available shapes, see VM Shapes |
| psft_ps_instance_listen_port_range | Port on which Peoplesoft process scheduler instance will listen and receive requests. |
| psft_tls_instance_shape | Shape of Peoplesoft tools instance. For more information on available shapes, see VM Shapes |
| psft_tls_instance_listen_port_range | Port on which Peoplesoft tools instance will listen and receive requests. |
| psft_web_instance_count | Number of Peoplesoft web server instances to be created. |
| psft_web_instance_shape | Shape of Peoplesoft web servers instances. For more information on available shapes, see VM Shapes |
| psft_web_instance_listen_port | Port on which Peoplesoft web server instance will listen and receive requests from Load Balancer. |
| psft_stage_filesystem_path | Mount path for Peoplesoft stage filesystem to stage software binaries. For example /stage/software |
| psft_stage_filesystem_size_limit_in_gb | Upper limit for Peoplesoft stage filesystem. This value is just to set an upper size limit visible to Peoplesoft. It does not restrict storage size of Filesystem Storage Service. |
| db_edition | Edition of database. |
| db_license_model | Licensing model for database. |
| db_version | Version of database. |
| db_node_count | Number of database nodes. For single instance database, this parameter should be 1 and for Real Application Cluster Database, this parameter should be set to 2. |
| db_instance_shape | Shape of Database nodes. For RAC, the minimum required shape is VMStandard1.2. |
| db_name | Name of Database Container. |
| db_size_in_gb | Size of database in gb. For more information, see Oracle Cloud Infrastructure Images |
| db_admin_password | Database administration password (sys password). |
| db_characterset | Characterset of database. |
| db_nls_characterset | National Characterset of database. |
| db_pdb_name | Starter Pluggable database name. |
| load_balancer_hostname | Hostname of the load balancer. |
| load_balancer_shape | Shape of the load balancer. |
| load_balancer_listen_port | Listen port of the load balancer. |
# AD (Availability Domain to use for creating Peoplesoft infrastructure)
AD = ["1", "2"]
# CIDR block of VCN to be created
vcn_cidr = "172.16.0.0/16"
# DNS label of VCN to be created
vcn_dns_label = "psftvcn"
# Operating system version to be used for compute instances
linux_os_version = "6.10"
# Size of boot volume (in gb) of compute instances
compute_boot_volume_size_in_gb = "100"
# Size of block volume (in gb) of compute instances
compute_block_volume_size_in_gb = "100"
# Login user for compute instance
compute_instance_user = "opc"
# Login user for bastion host
bastion_user = "opc"
# Timezone of compute instance
timezone = "America/New_York"
#Environment prefix to define name of resources
psft_env_prefix = "psftprd"
# Number of application instances to be created
psft_app_instance_count = "2"
# Shape of app instance
psft_app_instance_shape = "VM.Standard2.1"
# Listen port range of the application instance
psft_app_instance_listen_port_range = ["9033", "9039"]
# Number of process scheduler instances to be created
psft_ps_instance_count = "2"
# Shape of process scheduler instance
psft_ps_instance_shape = "VM.Standard2.1"
# Listen port range of the process scheduler instance
psft_ps_instance_listen_port_range = ["2320", "2321"]
# Number of elastic search instances to be created
psft_es_instance_count = "2"
# Shape of elastic search instance
psft_es_instance_shape = "VM.Standard2.1"
# Listen port of the elastic search instance
psft_es_instance_listen_port = "9200"
# Shape of tools instance
psft_tls_instance_shape = "VM.Standard1.2"
# Listen port range of the tools instance
psft_tls_instance_listen_port_range = ["5985", "5986"]
# Number of web instances to be created
psft_web_instance_count = "2"
# Shape of web instance
psft_web_instance_shape = "VM.Standard2.1"
# Listen port of the web instance
psft_web_instance_listen_port = "8000"
# Mount path for software stage filesystem
psft_stage_filesystem_path = "/stage/software"
# Set software stage filesystem limit
psft_stage_filesystem_size_limit_in_gb = "1024"
# Datbase Edition
db_edition = "ENTERPRISE_EDITION_EXTREME_PERFORMANCE"
# Licensing model for database
db_license_model = "LICENSE_INCLUDED"
# Database version
db_version = "12.1.0.2"
# Number of database nodes
db_node_count = "2"
#Shape of Database nodes
db_instance_shape = "VM.Standard2.4"
#Database name
db_name = "PSFTCDB"
#Size of Database
db_size_in_gb = "256"
# Database administration (sys) password
db_admin_password = "<password>"
# Characterset of database
db_characterset = "AL32UTF8"
# National Characterset of database
db_nls_characterset = "AL16UTF16"
# Pluggable database name
db_pdb_name = "DUMMYPDB"
# Hostname of Load Balancer
load_balancer_hostname = "psft.example.com"
# Shape of Load Balancer
load_balancer_shape = "100Mbps"
#Listen port of load balancer
load_balancer_listen_port = "8000"-
The Terraform modules create File Storage service file system to stage peoplesoft binaries. The filesystem is mounted on all compute instances.
-
The Terraform modules creates private load balancers with backends containing peoplesoft web servers.
-
Separate pairs of SSH keys can be used for bastion host and rest of the compute infrastructure resources. It is also possible to use the same key. In that case, same key is required as input to instance and bastion instance variables in env-vars or env-vars.ps1 file.
For example,
### Public/private keys used on the instance export TF_VAR_ssh_public_key=/home/oracle/tf/<mykey.pub> export TF_VAR_ssh_private_key=/home/oracle/tf/<mykey.pem> ### Public/private keys used on the bastion instance export TF_VAR_bastion_ssh_public_key=/home/oracle/tf/<mykey.pub> export TF_VAR_bastion_ssh_private_key=/home/oracle/tf/<mykey.pem>For terraform installations on Unix systems, the private half of SSH key pairs should be in OpenSSH format. The instances in private subnet can be reached via SSH on port 22 or port 3389 (windows) by allowing agent forwarding in Putty and using Putty authentication tool like Pageant. Note that this does not require copying private SSH key for instances to bastion host.
-
The terraform modules ensure that same tier instances are deployed across different Fault Domains with an availability domain. Fault Domains protect against unexpected hardware failures and against planned outages due to compute hardware maintenance. For Real application clusters database, each node of cluster is deployed in a separate Fault domain by default.
-
The terraform modules expose timezone variable which can be used to define timezone for provisioned compute instances. The code uses cloud-init to do that. For database system, timezone has to be set manually using OS specific procedure. Follow OS specific documentation to do that.
-
The terraform version has been locked to 0.11.8 and Oracle Cloud Infrastructure provider version has been locked to 3.5.1 in provider.tf file. To use a version higher than these versions, change the values in the provider.tf file. The terraform modules may require changes for a successful run with a new terraform and Oracle Cloud Infrastructure provider version.
Following are sample inline code to install RPMs, configure NTP and mount file systems on application servers:
#cloud-config
timezone: "${timezone}"
packages:
- nfs-utils
- oracle-rdbms-server-12cR1-preinstall.x86_64
- glibc.i686
- libstdc++.i686
- firefox.x86_64
- nc
- samba
runcmd:
- mkdir -p ${fss_mount_path}
- sudo mount ${fss_mount_target_private_ip}:${fss_export_path} ${fss_mount_path}
- echo '${fss_mount_target_private_ip}:${fss_export_path} ${fss_mount_path} nfs tcp,vers=3' | sudo tee -a /etc/fstab
- mount -a -t nfs- Go to Peoplesoft directory
cd Peoplesoft
-
Update env-vars (or env-vars.ps1 for Windows) with the required information. The file contains definitions of environment variables for your Oracle Cloud Infrastructure tenancy.
-
Update terraform.tfvars with the inputs for the architecture that you want to build. A running sample of this file for multiple availability domain architecture is available in previous section. The contents of sample file can be copied to create a running inputs file. Update db_admin_password with actual password in terraform.tfvars file.
-
Initialize Terraform. This will also download the latest terraform oci provider.
$ terraform init
- Set environment variables by running source env-vars on your UNIX system or by running env-vars.ps1 on your Windows system.
$ source env-vars
- Run terraform apply to create the infrastructure:
$ terraform apply
When you’re prompted to confirm the action, enter yes.
When all components have been created, Terraform displays a completion message. For example: Apply complete! Resources: 81 added, 0 changed, 0 destroyed.
- If you want to delete the infrastructure, run:
$ terraform destroy
When you’re prompted to confirm the action, enter yes.
Copyright © 2018, Oracle and/or its affiliates. All rights reserved. The Universal Permissive License (UPL), Version 1.0 Please see LICENSE.md for full details