Custom storage models can be created subject to Google Cloud Persistent Disk restrictions.
The below steps must be executed in sequence for using custom model before starting the SASE installation.
-
Copy /opt/rsa/saTools/cloud/sase-deployment-models.yml and /opt/rsa/saTools/cloud/host-models.yml to /root/.sase/. Update /root/.sase/host-models.yml with the custom model specification and Save.
-
Update the model_name attribute value with the above custom model name in /root/.sase/sase-deployment-models.yml and Save.
-
Complete the SASE node deployment. Refer to SASE Installation Guide.
All the attributes must be assigned a valid value. It is recommended to append custom model definition to existing content in /root/.sase/host-models.yml. Refer to the SASE Installation Guide for details on the attribute description.
Sample Custom Model Definition for Decoder
The sample custom configuration assumes defining a model capable of 1gbps capture with 10 days of packet retention and 40 days of concentrator meta retention : c1r10m40 , the machine_type as n2-standard-32 (Refer to Google Cloud documentation for details on all the available VM types) and disk_type as pd-standard and the storage requirements of 98000 GB for packet db and 8000 GB for other databases (metadb, sessiondb and index).
Due to gcp restrictions on the disk size, the model requires two sets of disks. Following the naming convention described earlier, the two disk sets are named (disk_name) as decoder, decodersmall, decoder0 and decodersmall0.
Custom model definition for model_name: c1r10m40.
host-models.yml:
Partial Content of sase-deployment-models.yml updated with custom model_name Value:
Sample Custom Model Definition for Concentrator
The sample custom configuration model (c1r10m40) assumes the machine_type as n2-standard-32 (Refer to Google Cloud documentation for details on all the available VM types) and concentrator meta retention of 40 days. The concentrator disk_type is pd-standard, index disk_type is pd-balanced and the storage requirements for metadb, sessiondb, root is 24000 GB (for 40 days) and index database is 6000 GB (for 40 days).
Since these sizes fall with in the GCP limitations on per disk disk maximum size, only one set (default name) is needed: concentrator, index.
Custom model definition for model_name: c1r10m40
host-models.yml:
Partial Content of sase-deployment-models.yml updated with custom model_name Value: