Managing StorageGRID Objects

Question 1

What is the life cycle of an object?

Answer 1

Ingest (client application creates object)
Manage (ILM manages the object over time and protects from data loss)
Retrieve (Client application reads object from storage node)
Delete (Client deletes the object using removal command)

Question 2

What is an ILM Policy?

Answer 2

A set of prioritised ILM rules that determines how the Grid manages the data over time.
This is applied when the object is first ingested.

Example rule: Place one copy in DC 1 disk, Place one copy in DC 2 disk, keep a copy on DC2 disk for one year before moving to cold storage in the cloud

Question 3

What is the default ILM Policy Rule?

Answer 3

The default rule is to make one copy of the object at every site.

Question 4

What is a storage pool

Answer 4

A logical grouping of storage nodes

Question 5

What is storage grade?

Answer 5

Defines the The type of storage that is using in a storage pool

Question 6

What is a site?

Answer 6

Is the data centre where storage pools are located.

Question 7

What is a region?

Answer 7

A location where ILM rules can be placed based on the regions where S3 buckets are created.

Regions can be used to keep data close to a customer location if required to reduce latency, address any regulatory requriements and minimise cost (if available).

Question 8

What is the Cloud Storage Pool?

Answer 8

This is a pool that allows you to move objects to an external storage (cloud) outside your Grid system.

Question 9

What are ILM rules?

Answer 9

They define
The location of the data
The type of storage that is used
The type of loss protection that is applied to an object’s data
The number of copies made
The life space of the object (and changes to its location, Renton and loss protection over time).

Question 10

What are the two mechanisms ILM rules provide to protect from object data loss?

Answer 10

Replication
Erasure coding

Question 11

What is object ingest?

Answer 11

When the client creates the object and sends it to the grid system (though a put request) matches the object against the ILM rule.

The system will make a temporary copy else where if it is unable to initially complete the rules of the ILM

Question 12

What is replication used for

Answer 12

To protect the object by making a copy at another site that can be updated by the GRID

Question 13

What is Erasure coding used for?

Answer 13

Data loss, where an ILM rule create object parity items which are spread across the GRID to ensure protection (like raid).
There are factors to consider
Erasure code (data + parity).
Storage overhead (%)
Storage Node redundancy — the number of nodes that can be lost without impacting data.
Site redundancy — The number of storage grid sites that can be lost without impacting data retrieval.

Question 14

What is the Object retrieve?

Answer 14

The client application retrieving data from the GRID system using a http get request.
Usually from the local storage node. The storage node returns the data via the gateway node.

Question 15

What is object delete?

Answer 15

This is when a client sends a delete request Http Delete API to the grid system.
All copies of an object are removed from the grid system.
It does not happen immedately. The key is removed first, and later the system will purge the data and meta data.

Question 16

What is the different between a storage pool and a cloud storage pool

Answer 16

A storage pool will reside within one or more sites.
Whereas a cloud storage pool will be one location (azure or AWS).
An object will be moved from the storage pool to the cloud but the meta data is stored in the Grid.

Question 17

What is cross-grid replication?

Answer 17

This allows fro the use of 3-2 backup stragegy (keep 3 copies, two in the grid, and one in the cloud).
All data is replicated between the grids.
Replication can be singe or bi-directional.

Question 18

What is object protection

Answer 18

Object Versioning (S3 bucketing)
Object Encryption (using data at rest)
Object lifecycle management (when objects are expired in certain S3 buckets)
Object Locking (this allows for client applications to set retention settings - i.e. for secure worm objects)

Question 19

What are Platform Services

Answer 19

They can help implement a hybrid cloud strategy (for your environment).

E.G.
CloudMirror Replication (ingested objects immediately replicated to an external S3 bucket).
Event notification (send an SNS notification when specific object event types occur).
Search Integration (Send object metadata to an Elasticsearch cluster when an object is ingested or deleted or when its
metadata changes).

Question 20

ASSESSMENT QUESTION - Which grid node thyme manages an object’s data and metadata storage, including protection?

Answer 20

Storage Node

Question 21

ASSESSMENT: Which grid node type is responsible for sending out AutoSupport information to NetAPP Support?

Answer 21

Admin

Question 22

ASSESSMENT: what is the minimum number of storage nodes that is supported at each site?

Answer 22

Three

Question 23

ASSESSMENT: A StorageGRID SG1000 appliance functions as which two types of grid node?

Answer 23

Admin
Gateway

Question 24

ASSESSMENT: Which management interface do you use to create a tenant?

Answer 24

Grid Manager interface

Question 25

ASSESSMENT: which two server types can be used as a federated identity source

Answer 25

AD
LDAP

Question 26

ASSESSMENT: A Tenant admin users which three items to identify and log in to a tenant account using the Tenant Manager Interface?

Answer 26

User name
Password
Account ID

Question 27

ASSESSMENT: Which to items are storage node attributes?

Answer 27

Site
??

Question 28

ASSESSMENT: Which three methods can be used to protect objects against site or node failures in the StorageGRID system?

Answer 28

Mirroring
Replication
Erasure coding

Question 29

ASSESSMENT: What is the default information lifecycle management (ILM) rule when the strorageGRID system is first installed?

Answer 29

Make a copy at each site