This template follows the documented steps https://cloud.google.com/solutions/sap/docs/certifications-sap-hana and deploys GCP and Pacemaker resources up to the installation of SAP's central services.
Install Terraform on the machine you would like to use to deploy from by following https://learn.hashicorp.com/tutorials/terraform/install-cli?in=terraform/gcp-get-started#install-terraform
-
Download .tf file into an empty directory
curl https://storage.googleapis.com/cloudsapdeploy/terraform/latest/terraform/sap_hana_scaleout/terraform/sap_hana_scaleout.tf -o sap_hana_scaleout.tf -
Fill in mandatory variables and if the desired optional variable in the .tf file.
-
Deploy
- Run
terraform init(only needed once) - Run
terraform planto see what is going to be deployed. Verify if names, zones, sizes, etc. are as desired. - Run
terrafom applyto deploy the resources - Run
terrafom destroyto remove the resources
- Run
-
Continue installation of SAP software and setup of remaining cluster resources as per documentation at https://cloud.google.com/solutions/sap/docs/sap-hana-ha-scaleout-tf-deployment-guide
For additional information see https://www.terraform.io/docs/index.html and https://cloud.google.com/docs/terraform
| Name | Description | Type | Default | Required |
|---|---|---|---|---|
| can_ip_forward | Whether sending and receiving of packets with non-matching source or destination IPs is allowed. | bool |
true |
no |
| custom_primary_metadata | Optional - default is empty. Custom metadata to be added to the primary VM. | map(string) |
{} |
no |
| custom_secondary_metadata | Optional - default is empty. Custom metadata to be added to the secondary VMs. | map(string) |
{} |
no |
| custom_standby_metadata | Optional - default is empty. Custom metadata to be added to the standby VMs. | map(string) |
{} |
no |
| data_disk_iops_override | Warning, do not use unless instructed or aware of the implications of using this setting. Directly sets the number of IOPS that the data disk(s) will use. Has no effect if not using a disk type that supports it. | number |
null |
no |
| data_disk_size_override | Warning, do not use unless instructed or aware of the implications of using this setting. Overrides the default size for the data disk(s), that is based off of the machine_type. | number |
null |
no |
| data_disk_throughput_override | Warning, do not use unless instructed or aware of the implications of using this setting. Directly sets the throughput in MB/s that the data disk(s) will use. Has no effect if not using a disk type that supports it. | number |
null |
no |
| data_disk_type_override | Warning, do not use unless instructed or aware of the implications of using this setting. Override the 'default_disk_type' for the data disk. | string |
"" |
no |
| disk_type | Optional - The default disk type to use on all disks deployed. The default is pd-ssd, except for machines that do not support PD, in which case the default is hyperdisk-extreme. Not all disk are supported on all machine types - see https://cloud.google.com/compute/docs/disks/ for details. | string |
"" |
no |
| hyperdisk_balanced_iops_default | Optional - default is 3000. Number of IOPS that is set for each disk of type Hyperdisk-balanced (except for boot disk). | number |
3000 |
no |
| hyperdisk_balanced_throughput_default | Optional - default is 750. Throughput in MB/s that is set for each disk of type Hyperdisk-balanced (except for boot disk). | number |
750 |
no |
| instance_name | Hostname of the GCE instance. | string |
n/a | yes |
| linux_image | Linux image name to use. | string |
n/a | yes |
| linux_image_project | The project which the Linux image belongs to. | string |
n/a | yes |
| log_disk_iops_override | Warning, do not use unless instructed or aware of the implications of using this setting. Directly sets the number of IOPS that the log disk(s) will use. Has no effect if not using a disk type that supports it. | number |
null |
no |
| log_disk_size_override | Warning, do not use unless instructed or aware of the implications of using this setting. Overrides the default size for the log disk(s), that is based off of the machine_type. | number |
null |
no |
| log_disk_throughput_override | Warning, do not use unless instructed or aware of the implications of using this setting. Directly sets the throughput in MB/s that the log disk(s) will use. Has no effect if not using a disk type that supports it. | number |
null |
no |
| log_disk_type_override | Warning, do not use unless instructed or aware of the implications of using this setting. Override the 'default_disk_type' for the log disk. | string |
"" |
no |
| machine_type | Machine type for the instances. | string |
n/a | yes |
| network_tags | OPTIONAL - Network tags can be associated to your instance on deployment. This can be used for firewalling or routing purposes. | list(string) |
[] |
no |
| nic_type | Optional - This value determines the type of NIC to use, valid options are GVNIC and VIRTIO_NET. If choosing GVNIC make sure that it is supported by your OS choice here https://cloud.google.com/compute/docs/images/os-details#networking. | string |
"" |
no |
| post_deployment_script | OPTIONAL - gs:// or https:// location of a script to execute on the created VM's post deployment. | string |
"" |
no |
| primary_startup_url | Startup script to be executed when the VM boots, should not be overridden. | string |
`"curl -s https://storage.googleapis.com/cloudsapdeploy/terraform/latest/terraform/sap_hana_scaleout/hana_scaleout_startup.sh | bash -s https://storage.googleapis.com/cloudsapdeploy/terraform/latest/terraform"` |
| project_id | Project id where the instances will be created. | string |
n/a | yes |
| public_ip | OPTIONAL - Defines whether a public IP address should be added to your VM. By default this is set to Yes. Note that if you set this to No without appropriate network nat and tags in place, there will be no route to the internet and thus the installation will fail. | bool |
true |
no |
| reservation_name | Use a reservation specified by RESERVATION_NAME. By default ANY_RESERVATION is used when this variable is empty. In order for a reservation to be used it must be created with the "Select specific reservation" selected (specificReservationRequired set to true) Be sure to create your reservation with the correct Min CPU Platform for the following instance types: n1-highmem-32 : Intel Broadwell n1-highmem-64 : Intel Broadwell n1-highmem-96 : Intel Skylake n1-megamem-96 : Intel Skylake m1-megamem-96 : Intel Skylake All other instance types can have automatic Min CPU Platform" |
string |
"" |
no |
| sap_deployment_debug | OPTIONAL - If this value is set to true, the deployment will generates verbose deployment logs. Only turn this setting on if a Google support engineer asks you to enable debugging. | bool |
false |
no |
| sap_hana_backup_nfs | Google Filestore share for /hanabackup | string |
n/a | yes |
| sap_hana_deployment_bucket | The Cloud Storage path that contains the SAP HANA media, do not include gs://. If this is not defined, the GCE instance will be provisioned without SAP HANA installed. | string |
"" |
no |
| sap_hana_instance_number | The SAP instance number. If this is not defined, the GCE instance will be provisioned without SAP HANA installed. | number |
0 |
no |
| sap_hana_sapsys_gid | The Linux GID of the SAPSYS group. By default this is set to 79 | number |
79 |
no |
| sap_hana_shared_nfs | Google Filestore share for /hana/shared | string |
n/a | yes |
| sap_hana_sid | The SAP HANA SID. SID must adhere to SAP standard (Three letters or numbers and start with a letter) | string |
n/a | yes |
| sap_hana_sidadm_password | The linux sidadm login password. If this is not defined, the GCE instance will be provisioned without SAP HANA installed. Minimum requirement is 8 characters. | string |
"" |
no |
| sap_hana_sidadm_password_secret | The secret key used to retrieve the linux sidadm login from Secret Manager (https://cloud.google.com/secret-manager). The Secret Manager password will overwrite the clear text password from sap_hana_sidadm_password if both are set. | string |
"" |
no |
| sap_hana_sidadm_uid | The Linux UID of the adm user. By default this is set to 900 to avoid conflicting with other OS users. | number |
900 |
no |
| sap_hana_standby_nodes | Number of standby nodes to create. | number |
1 |
no |
| sap_hana_system_password | The SAP HANA SYSTEM password. If this is not defined, the GCE instance will be provisioned without SAP HANA installed. Minimum requirement is 8 characters with at least 1 number. | string |
"" |
no |
| sap_hana_system_password_secret | The secret key used to retrieve the SAP HANA SYSTEM login from Secret Manager (https://cloud.google.com/secret-manager). The Secret Manager password will overwrite the clear text password from sap_hana_system_password if both are set. | string |
"" |
no |
| sap_hana_worker_nodes | Number of worker nodes to create. This is in addition to the primary node. |
number |
1 |
no |
| secondary_startup_url | DO NOT USE | string |
`"curl -s https://storage.googleapis.com/cloudsapdeploy/terraform/latest/terraform/sap_hana_scaleout/hana_scaleout_startup_secondary.sh | bash -s https://storage.googleapis.com/cloudsapdeploy/terraform/latest/terraform"` |
| service_account | OPTIONAL - Ability to define a custom service account instead of using the default project service account. | string |
"" |
no |
| standby_static_ips | Optional - Defines internal static IP addresses for the standby nodes. | list(string) |
[] |
no |
| subnetwork | The sub network to deploy the instance in. | string |
n/a | yes |
| unified_disk_iops_override | Warning, do not use unless instructed or aware of the implications of using this setting. Directly sets the number of IOPS that the primary's unified disk will use. Has no effect if not using a disk type that supports it. | number |
null |
no |
| unified_disk_size_override | Warning, do not use unless instructed or aware of the implications of using this setting. Overrides the default size for the primary's unified disk, that is based off of the machine_type. | number |
null |
no |
| unified_disk_throughput_override | Warning, do not use unless instructed or aware of the implications of using this setting. Directly sets the throughput in MB/s that the primary's unified disk will use. Has no effect if not using a disk type that supports it. | number |
null |
no |
| use_single_data_log_disk | Optional - By default two separate disk for data and logs will be made. If set to true, one disk will be used instead. | bool |
false |
no |
| vm_static_ip | Optional - Defines an internal static IP for the VM. | string |
"" |
no |
| worker_static_ips | Optional - Defines internal static IP addresses for the worker nodes. | list(string) |
[] |
no |
| zone | Zone where the instances will be created. | string |
n/a | yes |
| Name | Description |
|---|---|
| hana_scaleout_standby_self_links | List of self-links for the hana scaleout standbys created |
| hana_scaleout_worker_self_links | List of self-links for the hana scaleout workers created |
| sap_hana_primary_self_link | Self-link for the primary SAP HANA Scalout instance created. |
These sections describe requirements for using this module.
The following dependencies must be available:
- Terraform v0.13
- Terraform Provider for GCP plugin v4.0
Refer to the contribution guidelines for information on contributing to this module.
Please see our security disclosure process.