Deploying Kyvos GCP cluster with no-Spark through Kyvos Manager
Applies to: Kyvos Enterprise Kyvos Cloud (SaaS on AWS) Kyvos AWS Marketplace
Kyvos Azure Marketplace  Kyvos GCP Marketplace Kyvos Single Node Installation (Kyvos SNI)
Deploying Kyvos Manager
To configure the Kyvos GCP cluster in Kyvos Manager, you need to first deploy Kyvos Manager. For this, perform the following steps.Â
SSH to the node, where you want to install the Kyvos Manager.
Go to /data/kyvos/installs/ location.Download the Kyvos bundle file provided by the Support team using the following command.
wget <http://<BundleURL>
Where the BundleURL is the URL provided by the Kyvos Support team for downloading the Kyvos Bundle file.Untar the bundle file using the following command.
tar -xvf <bundle name>.tar.gz
Where bundle name is the name of your tar file, such as KyvosManager2020.4_ux64.Set the following parameters in the setenv.sh file present in the /kyvosmanager_war/kyvosmanager/bin folder.
KM_BASE_PATH=<install location>/kyvosmanager_war/
Note
Here, replace the <install location> with the path for Kyvos setup i.e. /data/kyvos/installs/<Kyvos version>
Go to /data/kyvos/installs/kyvosmanager_war/kyvosmanager/bin folder, and start the Kyvos Manager using the following command.
./start-km.sh
Deploying Kyvos GCP cluster with no-Spark
To deploy the Kyvos GCP cluster, perform the following steps.
Open the Kyvos Manager URL (such as http://<ip>:8080/kyvosmanager) on Google Chrome and log into it using administrator credentials.
Change the password after logging in, and re-login using the new password.
Now, navigate to http://<IP>:8080/kyvosmanager/kyvos-installer/gcp URL to start configuring the Kyvos cluster.
On the displayed wizard, select the I have all the resources option and click Install Kyvos. The Setup Kyvos environment dialog box is displayed.
On the Setup Kyvos Environment page, enter details as:
Parameter | Description |
---|---|
Cluster Name | Provide a name for your cluster. |
License | Click the Choose File button to select the license file and then click Upload. |
Installation Path | Provide the installation path to deploy Kyvos. |
User Name | Username with sufficient rights to deploy cluster. |
Authentication Type |
|
Select Instance by |
|
Project | Select your GCP project where your resources/deployment is present. |
Deployment | Select the None option (node mapping manually). Click the Fetch Resources link to get a list of resources available for your user credentials. |
Deploy BI Server, Query Engine, and Web Portal Kyvos Services on <IP address> instances | This checkbox is displayed only when you select the None (node mapping Manually)  option. NOTE: By default, the option to deploy instances for BI Server, instances for Query Engine, and instances for Web Portal will be disabled. |
Instance(s) for BI Server | Select the instances that you want to use for the BI Server. Optionally, you can also search the instance with its name. |
Instance(s) for Query Engine | Select the instances that you want to use for the Query Engines. Optionally, you can also search the instance with its name. |
Instance(s) for Web Portal | Select the instances that you want to use for the Web Portal. Optionally, you can also search the instance with its name. |
Repository | Select the instances that you want to use for the Repository. Optionally, you can also search the instance with its name. |
Repository Password | Provide the password for repository authentication. |
Secrets Manager |
NOTE: This option is available only for account-based resources deployments, which means if you selected the None option in the Deployment list and click the Fetch resources link then secrets list will be based on the currently selected project. To use different project, you can change the project for secrets. |
Output Storage Bucket Name | Provide the name of the bucket where Kyvos semantic models will be stored. |
SSH Preference | Select the checkbox to use SSH for connecting with Dataproc Cluster Master Node. |
User Name | User having requisite permissions for cluster deployment. |
Use different private key for Dataproc nodes authentication | Select this option to use a different private key for nodes' authentication. |
Compute Cluster Type | Select Shared Query Engine or Kubernetes from the list. NOTE: When you select Kubernetes from the Compute Cluster Type list, the Kubernetes Details section is displayed. |
Kubernetes Details | Enter the following Kubernetes details:
|
Enable Compute Server Scaling | Select this checkbox to enable maximum or minimum compute server for scaling. When you select this checkbox, the following options will be displayed.
NOTE: If you do not select this checkbox, the Compute Server Count field is displayed. Compute server scaling will be disabled if this checkbox is not selected. |
Function Deployment | Provide the name of the deployment where your Google Deployment Functions are deployed. |
Click the Advanced Settings link. The Advanced Settings dialog box is displayed with values populated according to the ones that you defined while creating the stack.
Configuration Type | Parameter/Field | Comments/Description |
---|---|---|
Common | Kyvos Setup Package | Select the Kyvos bundle to be installed. Kyvos Manager contains compatible Kyvos bundles. To use any other Kyvos version, click the Upload button |
Semantic Model Local Path | Enter the semantic model local path where cuboids will be copied. For multiple paths, enter a semicolon-separated list. | |
Automatically restart Kyvos services when down | Select this option to enable high availability mode for the Kyvos service. This will ensure Kyvos services are always up and running. The Kyvos Manager will automatically restart the BI server, query engine, repository, and web portal services, in case any of these services go down | |
BI Server   | Listener Port | Enter the listener port for BI Server. |
JMX Port | Enter the JMX port for BI Server health monitoring. | |
Work Directory | Enter the directory location on HDFS, where BI Server has access permissions. | |
Maximum Heap Memory | Enter the maximum memory available for BI Server. | |
Additional Java Options | Use this to define custom Java properties for your BI Server. | |
Query Engine  | Thrift Listener Port | Enter the thrift listener port for Query Engine. |
JMX Port | Enter the port for Query Engine health monitoring. | |
Maximum Heap Memory | Enter the maximum memory for Query Engine. | |
Additional Java Options | Use this to define custom Java properties for your Query Engine. | |
Web Portal    | UI Port | Enter the port on which Kyvos Web client will run. |
JMX Port | Enter the port for Web portal health monitoring. | |
Shut-down Port | Enter the port to be used for shutting the Tomcat server. | |
AJP Port | Enter the port on which the AJP connector creates a socket and awaits an incoming connection. | |
Maximum Heap Memory | Enter the maximum memory for the Web portal. | |
Additional Java Options | Use this to define custom Java properties for the Web portal. | |
Network Protocol Configuration | Enable TLS | Select the checkbox to enable TLS for the cluster. |
HTTP Protocol Version | Select the HTTP Protocol Version for use with TLS. | |
Kyvos Web Portal on HTTPS | Select the checkbox to use HTTPS for connecting to Kyvos Web Portal. | |
Kyvos Web Portal Configuration | NOTE: This is displayed only if Kyvos Web Portal on HTTPS checkbox is selected
|
Click Apply.
Click Next. The Summary page is displayed.
Click Install to complete the installation and deployment of the Kyvos GCP Cluster.
Copyright Kyvos, Inc. All rights reserved.