Getting Started¶

This chapter contains step-by-step instructions to create a VPSA and then to configure its storage properties.

Registering a Zadara Account & Creating a VPSA¶

Go to your organizational/cloud service provider VPSA Provisioning Portal
Click on Create account for first time user registration

Complete the form to create a Zadara Account.
go to your VPSA Provisioning Portal.
Login using your username / email and password, then press Create Zadara Storage VPSA® to create a new VPSA.

Note

Zadara has a public cloud Provisioning portal. User registration can be performed at: https://manage.zadarastorage.com/register/ .

The following dialog will appear:

Select either Storage Array to create a hybrid storage array or All Flash Array to create a performance optimized all flash array with built in inline data reduction support.

Note

VPSA Object Storage (ZIOS) creation is described in the VPSA Object Storage User Guide.

The VPSA definitions dialog will appear

Enter the following mandatory fields:

VPSA Name – Give the VPSA a name. This is how it will appear in the Cloud Console and in the VPSA GUI. If you are planning on having multiple VPSAs, you might want to give it as detailed a name as possible.
VPSA Description – Give the VPSA a description.
Select Cloud Provider – Select the Cloud or Co-lo where you have your compute instances. VPSAs are able to simultaneously connect to multiple Cloud Providers and Co-locations (within the same geographical region).

Note

From the public cloud Provisioning Portal you can provision and manage all of your VPSAs, even if they are connected to different Cloud Providers & regions.
Select a Region – Select the Cloud Provider region where your application servers reside. The servers and the VPSA must reside in the same region in order to establish efficient iSCSI or NFS\CIFS connectivity. Available Regions depend on which Cloud Provider you select.
Protection Zone – VPSA supports multiple protection zones in “stretched cluster” configurations, where each VC is in a different zone. In cloud locations that provide protection zones, select in which zone the new VPSA will be built. For multi-zone configurations select Multiple.

Click Next

Select the Zadara IO Engine – The Zadara IO Engine type defines the compute characteristics of your VPSA’s Virtual Controllers (VCs). Each engine type defines the following characteristics:
- Number of CPUs that are assigned to your VPSA’s VCs.
- Amount of RAM that is assigned to your VPSA’s VCs.
- Default size of protected SSD Cache.
When selecting the IO engine, take into account the capacity planned for this VPSA. Each Engine has a limit to the number of drive it can support, and to the total raw capacity of the VPSA.

You can change the Zadara Engine type (upgrade or downgrade) at any time throughout the lifetime of your VPSA according to your application’s needs,providing you stay within the maximum limits of the engine type you are moving to

Note

The compute resources (CPU, RAM and Cache) are dedicated to your VPSA, which ensures consistent performance and isolation from other tenants’ workloads and behavior.
Select the Cache size (for Storage Array engines larger than 200) Use the slide bar to set the amount of flash cache allocated to this VPSA. Note that each VPSA engine comes with a minimum amount of cache. The extended cache is allocated in 200GB increments.

Drive Quantities – Select the type and number of drives that you would like allocated to your VPSA.

The Zadara Cloud Orchestrator allocates dedicated drives.
Drives are allocated from as many different SNs as possible to provide max redundancy for your VPSA’s RAID groups.

There is a limit to the number of drives per Zadara IO Engine type. The larger the Engine is, the more drives you can add. There is also a limit to the total raw capacity of all drives. Make sure the total capacity of all selected drives is within the limit.

The following table lists the maximum drives per Storage Array Engine type:

IO Engine Type	Maximum # of Drives	Maximum Raw Capacity
200 (Baby)	5	24 TB
400 (Basic)	10	60 TB
600 (Boost)	20	100 TB
800 (Blast)	30	150 TB
1000 (Blazing)	40	200 TB
1200	60	240 TB
1600	80	300 TB
2400	80	360 TB
3600	80	360 TB

The following table lists the maximum drives and capacity per All Flash Array Engine type: Note that for All Flash Array, due to data reduction, the capacity limit per engine depends on both the physical capacity of the drives and the the customer virtual capacity (as seen by the hosts), before any data reduction. More about All Flash Array capacities: Understanding Pool’s Capacity

IO Engine Type	Maximum # of Drives	Maximum Raw Capacity	Maximum Customer (host) Capacity
F800	30	120 TB	60 TB
F1200	60	200 TB	100 TB
F2400	80	320 TB	160 TB
F3600	100	400 TB	200 TB
F4800	120	400 TB	250 TB

Note

The above capacities depend on the type of the pool(s) used. The numbers shown are the limits of the aggregated size of all pools of type Archival. See Creating a Pool for details

Click Next

Select the Zadara Container Services (ZCS) Engine – The Zadara ZCS Engine defines the compute resources of the VPSA’s Virtual Controllers that are allocated for Docker containers within this VPSA. Refer to Managing Container Services for details about Zadara Container Services.
Fibre Channel Support – Check this checkbox if you will be connecting hosts to this VPSA over FC SAN.

Click Next

Create RAID-10 Pool(s) – By default, at VPSA creation time RAID-10 pools are automatically created. One pool per each drives type selected. All the drives selected Of each type will be included in the pool. If you want to create different pools setting, uncheck this checkbox, and manually create your RAID groups and pools as described below.

Once you have completed selecting the above VPSA characteristics, review the displayed summary. You can click Edit to modify your previous selections. Press the Create button to confirm the VPSA creation request. The requested VPSA will appear in the “Awaiting Approval” list.

Completing the VPSA creation requires the approval of a Zadara Storage Cloud admin. Once approved, the new VPSA only takes a few minutes to launch. During that time you’ll see your VPSA in “Launching” status as shown below:

Once the VPSA is ready, you’ll receive an email with a temporary passcode at your registered email address.
Use the “Management Address” link to access the VPSA GUI:

Use your registered username or email address and the temporary passcode to enter the VPSA GUI. You will be immediately prompted to set a new password for your VPSA User account.

The VPSA Interface¶

Understanding the VPSA Dashboard¶

VPSA Dashboard is the landing page, every time the GUI opens. It gives the overall state of the VPSA (Health, Capacity, Performance) at a glance. The Dashboard is made of the following components:

VPSA Info: General information of the VPSA such as name, engine type and management IP Address.
System Health: Shows the inventory of the objects managed by the system, such as Pools, Volumes, Mirrors etc… If all objects are in “normal” state, there is a green checkmark on the line. If there is situation that needs your attention a red exclamation mark is shown on that specific line, with number of objects that need to be taken care of.
CPU: Shows the CPU utilization of the active Controller of the VPSA, over time. This chart gives an indication of the load on the storage system.
Capacity: Shows the capacity state of the VPSA. The display is different between Storage Array and All Flash Array. For the later it shows the capacity reduction saving. See Understanding Pool’s Capacity for details.
1. Current capacity state
2. Capacity consumption over time during the last month
Performance: These charts show the aggregated performance of all Volumes.
1. Current IOPS (reads and writes) of all Volumes
2. IOPS activity during the last hour
3. Current throughput of all Volumes
4. Throughput of all volumes during the last hour
5. Current mirroring traffic of all mirrors (outbound and inbound)
6. Mirroring activity of all mirrors during the last hour

Understanding the VPSA GUI¶

The VPSA GUI provides full management and control of your VPSA. It contains the following main components (as numbered in the above screenshot):

Main Navigation Left Panel – Traverse through the various VPSA entities. The selected entity is highlighted.
The Center Pane – Displays a list of objects from the selected entity type (e.g. Drives in the above screenshot example) and for each object it displays the main properties.
The South Pane – Displays detailed information regarding the selected object. All objects have at least 3 tabs:
1. Properties – Detailed properties of the object.
2. Metering – Typically IO workload metering info.
3. Logs – List of event-log messages related to that object.
Logged-in username – Displayed at the top right corner.
Selected Language – Displayed at the top left corner. You can use this drop down to change the displayed language.

Creating RAID Group, Pools, and Volumes¶

By default a new VPSA is created with all its drives configured in RAID Groups, and a Pool per each drives type. If the automatic pools satisfy your needs go directly to the volumes creation below. Otherwise follow the RAID Group and Pool creation instruction.

Create a RAID Group to define the level of data protection needed. For more details see here: Creating a RAID Group
Create a storage Pool by using aggregated capacity from one or more RAID Groups. For more details see here: Creating a Pool
Create an iSCSI\FC\NFS\SMB Thin Provisioned Volume to be used by your servers. For more details check here: Creating and Deleting a Volume
Add a server. The server object represents the host using the storage volume. Follow the instructions depending on OS and connectivity of your server: Adding a Server
Attach the Volume to a Server. For more details see here: Attaching & detaching Volumes to Servers

Congratulations! You have a new VPSA provisioned and ready to use.

The following sections describe in detail the various capabilities and services of your VPSA.

Provisioning your VPSA¶

You create, add, change, delete and manage the resources composing your VPSAs via the Zadara Provisioning Portal.

This section describes the available operations in the Provisioning Portal (https://manage.zadarastorage.com).

Adding and removing Disk Drives¶

To add drives to your VPSA go to the Provisioning Portal, select the VPSA and then press the Add Storage button.

Select the number of drives from each available drive type you wish to add to your VPSA, and press Submit. Keep in mind the RAID Groups you are going to build.
This operation requires the approval of a Zadara Storage Cloud Admin. Once approved, you’ll see the number of drives in the Provisioning Portal update accordingly. If you then refresh the Drives page in the VPSA GUI the new drives will be displayed.

You can remove unused Drives (indicated with status “Available”) from within the VPSA.

Go to the VPSA GUI > Drives, select the Drive you wish to remove and press the Remove button:

If you wish to remove a Drive that is part of a RAID Group you first need to replace it with another Drive as described here: Replacing a Drive

Managing Zadara Engines¶

The Zadara IO Engine type defines the following characteristics of your VPSA’s Storage Controllers:

Dedicated CPU and memory resources - These are dedicated solely to your VPSA. These resources are not shared with any other VPSA or tenant within the Zadara Storage Cloud.
Flash Cache Size - Each VPSA is provisioned with a Flash Cache partition to be used for both metadata and read/write caching. The SSD cache partition is protected using RAID-1, where each mirror copy resides on a different SN, thus ensuring cache resilience to SN failure. Each Engine type comes with a base SSD cache partition size. You can request additional flash capacity for caching. For more details see “Adjusting Cache.”
Maximum number of drives – The maximum number of drives that can be allocated to each VPSA engine type.

The following Zadara IO Engines are available For Storage Array:

IO Engine Type	Dedicated Compute Resources	Base Flash Cache	Max # of Drives	Max Raw Capacity
200 (Baby)	2 CPU, 6 GB RAM	20 GB	5	24 TB
400 (Basic)	4 CPU, 12 GB RAM	20 GB	10	60 TB
600 (Boost)	6 CPU, 20 GB RAM	40 GB	20	100 TB
800 (Blast)	8 CPU, 28 GB RAM	60 GB	30	150 TB
1000 (Blazing)	10 CPU, 36 GB RAM	80 GB	40	200 TB
1200	12 CPU, 52 GB RAM	100 GB	60	240 TB
1600	16 CPU, 68 GB RAM	120 GB	80	300 TB
2400	24 CPU, 100 GB RAM	180 GB	80	360 TB
3600	36 CPU, 144 GB RAM	240 GB	80	360 TB

The following Zadara IO Engines are available For All Flash Array: Note that for All Flash Array, due to data reduction, the capacity limit per engine depends on both the physical capacity of the drives and the the customer virtual capacity (as seen by the hosts), before any data reduction. More about All Flash Array capacities: Understanding Pool’s Capacity

IO Engine Type	Dedicated Compute Resources	Maximum # of Drives	Maximum Raw Capacity	Maximum Customer (Host) Capacity
F800	8 CPU, 48 GB RAM	30	120	60 TB
F1200	12 CPU, 72 GB RAM	60	200	100 TB
F2400	24 CPU, 116 GB RAM	80	320	160 TB
F3600	36 CPU, 176 GB RAM	100	400	200 TB
F4800	48 CPU, 236 GB RAM	120	400	200 TB

Note

The above capacities depend on the type of the pool(s) used. The numbers shown are the limits of the aggregated size of all pools of type Archival. See Creating a Pool for details

The following Zadara Container Services Engines (see: Managing Container Services) are available:

Zadara ZCS Engine Type	Dedicated compute resources
01	2 CPU, 512 MB RAM
02	2 CPU, 1 GB RAM
04	4 CPU, 2 GB RAM
06	6 CPU, 4 GB RAM
08	8 CPU, 8 GB RAM

To change both types of Zadara Engines, press the Change Engine link in the Zadara Provisioning Portal:

When selecting any engine larger than 200 you can also select the required flash cache size for that engine. For Flash Cache limits see here.

Completing this operation requires the approval of the Zadara Storage Cloud Admin.

The Zadara Engine upgrade/downgrade process may take a few minutes. During that time, your VPSA status will change to “Upgrade Pending”.

When the process completes the VPSA status will change back to “Ready”.

Managing Virtual Networks¶

The Zadara cloud provides a flexible and dynamic virtual networking infrastructure that can be tailored to meet multiple storage architecture and use cases.

Each cloud tennant is allocated with one or more “Virutal Networks” which is a set of available IP addresses within a specific network segment. Virtual networks are allocated for a specific cloud tenneant and within a specific avilable cloud VLAN.

The below diagram depicts the relationship between cloud tenants, virtual networks and VLANs:

In case a VPSA serves as storage for servers on different networks, the VPSA can be plugged on multiple “Virtual Networks”. Both block volumes and NAS shares can simultaneously be exposed through one or more Virtual Networks.

Each VPSA is created with a primary network for its front end (hosts connectivity). This network is routable and is mandatory.

You can manage your virtual networks from the Zadara Provisioning Portal. click Network Management in the Zadara Provisioning Portal’s Management Menu.

To create another virtual network press Create , and fill in the requested parameters such as: CIDR, Gateway, IP Range, and whether IPv4 or IPv6 should be used.

You can add (and remove) secondary networks to the VPSA. The VPSA internally maintains a “Virtual Networks Interface” (VNI) that connects into each virtual networks.

To add Virtual Network Interface, press the Add Virtual Network Interface link in the Zadara Provisioning Portal VPSA Operations.

To remove Virtual Network Interface, press the Remove Virtual Network Interface link in the Zadara Provisioning Portal VPSA Operations.

Note

Number of VNIs per VPSA is limited to 5.
VPSA REST API/GUI is accessible through any VNI.
Only Primary VN IP is registered in DNSimple
VPSA can’t have two VNI with the same VLAN.

Note

Only “Primary Virtual Network” is a routable network. Remaining virtual networks are not routable. There are some limitations on the remaining virtual networks:
Active Directory can be joined only through “primary virtual network”.
Backup (B2OS), Mirror, Remote Clone through FE network are only allowed via the “primary virtual network”.
ZCS container services exposed through FE network can be done only on “primary virtual network”.
“iSER” host connectivity is available only on the “Primary Virtual Network”.

Assigning Public IPs¶

By default you cannot access the VPSA from the public Internet for security and privacy reasons. The VPSA Front-End IP address which is used for VPSA management (via GUI and REST API) and for data IO workload (host connectivity via iSCSI/NFS/SMB protocols), is allocated on the Zadara Storage Cloud “Front-End” network 10GbE interface which is routable only from the Cloud Servers network. Servers outside of your Cloud Servers network cannot reach this IP address. This means you cannot access your VPSA GUI from the Internet.

A typical use case requiring Public IP addresses is when you’re running Asynchronous Remote Mirroring between two VPSAs in different regions, between on premise and cloud deployments or even between different Cloud Providers for Disaster Recovery (DR). Communication between the VPSAs is done via an authenticated and encrypted channel over the public Internet, thus requiring Public IPs.

To assign a Public IP address to your VPSA, go to the Provisioning Portal and press the Assign Public IP link. You can see the assigned IP address in your VPSA details in the Provisioning Portal and in the VPSA GUI, under Settings > General > Public IP. To remove it, simply click the Remove Public IP button in the Provisioning Portal.

Note

Access to the VPSA GUI and API is blocked through the Public IP for security reasons.

Note

NAT’d server IP connections are not supported for iSCSI, NFS, and SMB protocols over the Public IP.

Adjusting Flash Cache¶

Each VPSA is provisioned with a base flash cache partition, which is utilized by the VPSA for both metadata and read/write caching. The initially assigned default SSD cache size is also the minimal cache size for a given Zadara Engine. The flash cache partition is protected using RAID-1, where each mirror copy resides on a different SN, thus ensuring cache resilience to multiple types of failure.

On top of the base flash cache described above, you can add an extended cache. The VPSA extended flash cache size is elastic, so you can increase or decrease the cache size according to the needs of your workload.

Each Engine type has a minimum (default) and maximum SSD Cache size, as shown in the table below:

Zadara Engine	Base Flash Cache	Default Extended Flash Cache Size	Max Extended Flash Cache Size
200 (Baby)	20 GB	0 GB	0 GB
400 (Basic)	20 GB	200 GB	400 GB
600 (Boost)	40 GB	400 GB	800 GB
800 (Blast)	60 GB	600 GB	1200 GB
1000 (Blazing)	80 GB	800 GB	1600 GB
1200	100 GB	1200 GB	2400 GB
1600	120 GB	1600 GB	3200 GB
2400	180 GB	1600 GB	3200 GB
3600	240 GB	1600 GB	3200 GB

To change the Extended Flash Cache size for your VPSA, go to the Provisioning Portal and press the Adjust Flash Cache link:

Hibernating your VPSA¶

You can hibernate your VPSA when it is not in use for some period of time in order to reduce its associated service cost. While the VPSA is in a hibernated state you will only be billed for the drives, not the engine. Hibernating a VPSA involves the process of deleting its Virtual Controllers (the VPSA) while maintaining the data drives and all the necessary metadata to resume its operation at a later stage. No data is lost! The hibernated VPSA is not accessible to any GUI or REST API commands, nor will it present any iSCSI or NFS\SMB volumes. Resuming a hibernated VPSA only takes a few minutes.

To hibernate a VPSA, go to the VPSA Provisioning Portal and press the Hibernate link:

To resume access to the VPSA, go to the Provisioning Portal and press the Restore link. (The Hibernate and Restore toggle depending on the current state of the VPSA.)