Deploying Panzura as a Google Cloud Engine (GCE) Virtual Machine

The following deployment is only applicable for Panzura CloudFS Version 8.

GCE Prerequisites for node Deployment

You will need the following information when deploying the Panzura node on Google Compute Engine:

 

Item Your Value
Username  
Password  
Bucket name  
Region’s Hostname or IP Address  

If you don’t have an account, go to https://console.cloud.google.com/getting-started to create one.

Panzura node Information and Prerequisites

Completing the table below will reduce the time it takes to deploy the node. For nonMaster deployments, the licenses, cloud storage provider information, password and supported protocols are inherited from the master.

Panzura recommends using Chrome as your browser for both completing the setup wizard and when using the node WebUI.

Task Your Value Description
EULA Acceptance    
Name, Email address, Company Name, Title   There is an option to download the EULA.
New Password   Minimum 8 characters and should meet your corporate security requirements
Network Settings    
Shared Network / Dedicated Network  

Shared: single network interface used for client and cloud traffic.

Dedicated: separate network interface for client and cloud traffic.

DHCP / Static    
Shared Network - IP Address    
Shared Network - Subnet Mask    
Shared Network - Default Gateway    
Primary DNS Server IP    
Secondary DNS Server IP   Optional
Enable Jumbo Frame   On/Off. Contact your network administrator and Panzura Support before enabling Jumbo Frame support.
Dedicated Network - IP Address   If applicable, Cloud (WAN) Interface.
Dedicated Network - Subnet Mask   If applicable, Cloud (WAN) Interface.
Dedicated Network - Default Gateway   If applicable, Cloud (WAN) Interface.
Dedicated Network - Enable Jumbo Frame   On/Off. Contact your network administrator and Panzura Support before enabling Jumbo Frame support.
System Settings    
Hostname   Registered name of the node in the DNS.
node Geographic Location   Street address, City, State / Province, Country, Zip code / Pin code.
Contact Email Address    
DNS Domain    
Role Settings    
Configuration Mode   Default is Subordinate. If this is the first node, select Master. Options are Master / Subordinate / HA -Local / HA - Global.
Collaborative Mesh   Full Mesh or Hub & Spoke deployment architecture. Default is Full Mesh.
Private Secure Mode   Enable this feature if deploying in secure site with no public internet access.
Support Assistance   Enable this feature to allow Panzura Support to remotely provide assistance with troubleshooting, and to send daily logs to Panzura.
Configure as Key Master   Master node is the Authentication Key Master for peer -peer node communication. If the node is the Master, export (download) the key. For all other configurations, upload the pairing key exported from the Master.
Datastore Settings & Allocation    
Datastores: Automatic or Manual  

For Automatic, upload the json file with privileges for creating datastores. For Manual, datastores must be pre - allocated to the node. See Instructions in the Appendix.

Panzura recommends SSD storage or provisioned IOPS for metadata and local cache.

Cloud storage available    
Percentage of the total data accessed frequently   Percentage of the total data that is accessed frequently. This can be an estimate.
Panzura License  
Panzura License Token This is a hexadecimal string provided by Panzura containing purchased licenses. This can be skipped if you have license files. Contact sales@panzura.com for the license token.
Cloud Storage Provider  
Cloud storage provider (CSP) Panzura supports many public and private CSPs. Refer to your object store provider for the credentials you will need for the object store you plan to use.

 

GCE VM Instance Sizing

The following table lists the minimum GCE system requirements:

Size Speed Name Storage
Small 1 Gbps n1-standard-8 SSD, Provisioned IOPs
Medium 2 Gbps n1-standard-16 SSD, Provisioned IOPs
Large 4 Gbps n1-standard-32 SSD, Provisioned IOPs

Make sure to consult with you Panzura Sales Architect for specific VM sizing for your environment.

Create a Google Cloud Platform (GCP) Account

  1. 1. To create a Google Cloud Platform (GCP) account, open a browser and go to: https://cloud.google.com/
  2. Click Sign In as shown below:
  3. Create your account by providing your business email address and email:

 

We recommend creating a Google account specifically for use with your Google Cloud storage and Panzura deployment, and not one tied to any individual user, to prevent accidental deletion.

Deploying a Panzura node as a GCE VM

  1. In your Chrome browser, log in to Google Marketplace ( https://cloud.google.com):
  2. In the left navigation panel, select Marketplace:
  3. On the Marketplace page, search for Panzura and select the Panzura node offering. You can deploy a node using the Cloud Launcher. For this, simply go to https://console.cloud.google.com/marketplace/browse?q=panzura
  4. Click the “LAUNCH” button
  5. Under Deployment name, enter the name of your Virtual Machine
    1. Choose a Zone where your VM will be deployed.
    2. Consult with your Panzura Sales Architect for appropriate sizing (vCPU and memory).
    3. Choose “SSD Persistent Disk” as “Boot disk type” and set the “Boot disk size in GB” to 60.
    4. Select a Network Interface.
    5. Mark the check box which allows for TCP traffic from the internet. If you have further restrictions, make sure to limit the IP ranges according to your corporate policy.
    6. Mark the check box which allows for HTTPS traffic from the internet. If you have further restrictions, make sure to limit the IP ranges according to your corporate policy.
    When complete, click the “Deploy” button.
  6. When your node deployment is complete, you will receive node details.
  7. Please copy the Admin Password to use for the Panzura Setup Wizard. When you are done, click on the Admin URL

node Setup Wizard Steps

  1. Use a Chrome browser to connect to the Panzura GCE instance using the external IP address assigned to the node during instance deployment. Enter the IP address in the Chrome browser URL: https://node-ipaddr
    Since there is no certificate loaded for the node, the “Your connection is not private message” appears. Click Advanced, then click the “Proceed to nodeIPaddr (unsafe)” link to continue to the Panzura management WebUI.
  2. Log in with username admin and enter/paste the password visible on the Google Console under the Admin Password section. You can ignore the field above Username, this will display the node’s hostname once assigned.

It may take 5 minutes for the node to fully power on and respond to a web login request.

3. After you click the “Log In” button,the Configuration Wizard page appears. Click Start to begin the wizard setup:

 

4. Read and accept the Panzura End User License Agreement (EULA). You must click I ACCEPT to proceed or exit without proceeding. Include your name, email address, company name and title. Click Next.

5. Review the setup requirements, then click Next:

6. Change the default password for the admin account to a secure password that meets your corporate policy for creating secure passwords. Your new password must be at least 8 characters in length. You must then re-enter your password in the next field. You will not be able to proceed if they do not match. Remember your password, as you will not be able to change or reset your password in the future without it.

The Welcome icon is blue indicating this section is complete. The next section is the Network Settings for the node.

When you are ready, click the “Next” button:

7. Configure network settings.

    1. Network (Client & Cloud) Settings: If you have a single network interface for client (LAN) and cloud (WAN) traffic, select Shared Network. If the client and cloud traffic are on separate network interfaces, select Dedicated Network interface. This setting will display additional fields for you to enter the network information for cloud traffic.
    2. Client(LAN) Interface Settings: Select DHCP or Static. If you started off with DHCP but prefer to assign a static IP address, enter the following values:
      1. IP address
      2. Subnet mask
      3. Default gateway address
      4. Primary DNS server
      5. Secondary DNS server (optional)
    3. Jumbo Frames: This disabled by default. Consult your network administrator and Panzura Support prior to enabling this feature.In this example, Shared Network is selected. The address allocation method is left set to DHCP. (This combination is not required. For example, the Static option can be selected with either Shared Network or Dedicated Network.)
    4. Proxy Server: This option is only supported for use with F5 products and specific object stores. Please check the Admin Guide for supported configurations.

When you are ready, click the “Next” button:

8. Configure system settings.

    1. Hostname: This is the hostname of the node and should be unique and DNS resolvable.
    2. node Location: Enter a valid geographical address or latitude and longitude.
    3. Contact Email: Enter a valid email address.
    4. Domain Name: Enter the DNS domain to which this node belongs.

When you are ready, click the “Next” button:

9. Panzura Licensing. Enter either your Panzura provided token or License Files and click the “Next” button:

10. Time Settings. Choose whether you would like to use a Network Time Protocol (NTP) server. You can enter up to 4 NTP servers, separated by commas.

When you are ready, click the “Next” button:

11. Configure node role settings. This section of the wizard configures the following functions for the node:

    1. Role Settings:
      1. Configuration Mode: Select the node's role in the CloudFS:
        1. Master: This node is the first node you are setting up for the CloudFS.
        2. Subordinate: This node is an active node (not an HA node) that also is not the Master.
        3. HA-Local: This node will take over for a specific individual node if that node becomes unavailable.
        4. HA-Global: This node can take over for one of multiple nodes in the cluster that become unavailable.

If the node is a non-Master, the Master node Hostname field appears. You will need to enter the Master node’s hostname since the Subordinate will pull its configuration information from the Master. If Auto Failover is enabled, enter the shared hostname in this field.

ii. Collaborative Mesh: Select Full Mesh or Hub & Spoke. This option is available if the node is a Master.

iii. Configure as a Key Master: This enabled by default on the Master and is not available for non-Master configuration modes. At the bottom of the page, click Export to export the “pairing.key” file. When using the setup wizard to configure Subordinate or HA nodes, you will need to upload this key file onto those nodes. This will enable all the nodes in the cluster to securely communicate with each other using a uniquely specific authentication for your environment.

iv. Private Secure Site Mode: Enable this if there is no public internet access allowed for the node.

v. Support Assistance: Enable uploading of support logs to Panzura Support to remotely troubleshoot and resolve cases quickly. This is disabled if Private Secure Site Mode is enabled.

vi. Peer-to-Peer Authentication Key: If the node is a Master, export the key file and save it locally. An Upload button will appear for non-Master node configurations for you to upload the pairing.key file. This will enable the nodes to communicate with each other in the cluster. If it is a Master, there is an option to regenerate the key.

When you are ready, click the “Next” button: 

12. Configure datastore settings.

This section will enable you to discover datastores that will used to assign to the node. Panzura requires SSDs for metadata and cache storage, which can be assigned automatically or manually. The datastores must be in RAID groups to spread the load across multiple drives and preserve access to data in the event of drive failure.

a. Automatic: The wizard will discover the datastores to assign to the node by uploading a json file containing the private key to authorize the node to create datastores.

b. Manual: If the datastores have already been assigned to the node.

c. Cloud Storage To Allocate: Enter the estimated cloud storage. This is the amount of data that will be moved to cloud storage and managed by the node. This amount is the raw capacity of the node (data that is not compressed or deduplicated by the node).

d. Percentage of Cache: This is the estimated percentage of data that will be accessed frequently. A value of 10 (percent) is appropriate for most environments.

The Cloud Storage and Percentage of Cache can be expanded later without additional cost but does require a fully licensed node token.

When you are ready, click the “Next” button:

13. Configure datastore allocation settings.

This section of the wizard allows you to assign datastores to metadata and cache. Panzura recommends SSDs or datastores with SSD-like high performance random access for metadata and cache. Consult your Panzura Sales Architect for proper sizing for your environment. For each attached disk, select:

a. Media Type. This can be either HDD or SSD.

b. Datastore Type. This can be either Metadata or Cache. You must have at least one (1) drive for Metadata and one (1) drive for Cache.

You can create new disks (refer to Google Cloud Platform documentation for creating and attaching local disk to your VM) while in the node Setup Wizard. If you do so, you can click the “Discover New Disks” button to reveal them in the Datastore Allocation table.

When you are ready, click the “Next” button:

14. Once a disk has been allocated for metadata or cache with the Panzura node, it cannot be re-allocated.

When you are ready, click the “Proceed” button:

15. Configure Cloud Storage Provider (CSP) Enter the information for the cloud storage provider that this node will use to store the data in the cloud.

Refer to the Panzura Admin Guide for details on the available cloud providers and the information required for each vendor to connect to the storage. The table also lists the network ports to leave open. See the online help for additional details.

If you would prefer to configure the cloud storage provider settings later, select None.

The following example is for Google Cloud Storage (GCS). When you are ready, click the “Next” button:

Enter a desired path within the bucket. If this path does not already exist, Panzura node will create the path within the bucket and use this path to establish the file system. If the path does exist, this indicates that a file system already exists, and the Panzura node will use the path given.

For subordinate and HA nodes, the CSP information will be retrieved from the Master node except for the secret key or password for cloud storage.

16. Once you enter all the cloud credentials, you will see the following message to acknowledge before proceeding. When you are ready, click the “Proceed” button:

17.Protocol Settings: Select the protocols this node will support. Panzura supports SMB, NFS or both.

For non-Master deployments, protocol information is retrieved from the Master.

When you are ready, click the “Next” button:

18.Enter Active Directory (AD) domain settings:

a. AD Domain Name: Enter the Active Directory server name this node will use to authenticate users.

b. NetBIOS: Optional.

c. Domain Controller: Panzura recommends selecting “Any”in case the primary DNS fails.

d. AD Administrator, AD Administrator Password: Enter the AD administrator login and password to connect this node to the AD server.

When you are ready, click the “Next” button:

19.Review the settings and finish the wizard. You can select Back to make any changes or click Finish to complete the wizard.

20.Once the wizard completes the node configuration, you will have the option to go to the main WebUI home page:

21.If the node is deployed as the CloudFS Master, the following icons appear: