Module 4 Flashcards
Azure spans the globe with more than how many facilities
100
What is a Region in Azure and what does it contain.
- A region is a geographical area containing one or more data centers that are nearby and networked via low latency network.
- Azure assigns and controls the resources within a region to ensure workloads are appropriately balanced.
True or False, Everything Azure offers is available in every region
False, some things are only available in certain regions (i.e. VM Size or Storage Types)
What types of things do not require a region
- Some Things are global services that do not require region selection (i.e. Azure Ad, Traffic Manager and Azure DNS).
-Azure has more global regions than anyone else, giving the flexibility to get your App closer to your user wherever they are, as well as also providing what…
Better scalability, redundancy and better preserves data residency
Azure has some special regions, what are these and why do they exist
US DOD Central US Govt Virginia US Govt IOWA and More China East China North and More
These exist for compliance and legal reasons
How are the US Govt regions separated and how are they operated
- Physically and logically separated for US govt agencies and operated by screened US personnel
How are the China Regions managed
Managed via Partnership with 21vianet, MS does not directly manage these.
What are regions used to identify
Regions are used to identify the location of your resources
Azure divides the world into geographies, how are these defined
- Azure divides the world into geographies, defined by country borders or geopolitical boundaries
- Azure Geography is a discrete market containing 2 or more regions preserving data residency boundaries.
What do Geographies allow customers to achieve
- Geographies allow customers wit specific data residency and compliance needs to keep apps and data close
- Ensure data residency sovereignty, compliance and resiliency requirements are honoured within boundaries
What level of fault tolerance are built into geographies and how is this achieved
Fault Tolerant to withstand complete region failure through a connection to a dedicated high capacity network infrastructure.
What areas are geographies broken down into
- Americas
- Europe
- Aspac
- Middle East and Africa
In what way do regions belong to geographies and what is applied
Each region belongs to a single geography and has specific service availability, compliance and data residency applied.
You want to make your data and services redundant to protect in case of failure, when using On-Prem this requires hosting duplicate H/W environments.
How does Azure help with this
- Azure can help through availability zones
What are Availability Zones made up of
Availability Zones are physically separate DCs within an Azure region
- Each Availability zone is made up of one of more DCs with independent power/cooling/networking
- If one zone goes down the other continues working.
- Availability Zones are connected via High Speed fibre networks
Does every region support Availability zones
No, not every region support availability zones
How do Availability zones allow running of mission critical apps
- Availability Zones allow running of mission critical apps building High availability bu co-locating compute, storage, data resources within a zone and reduplicating it in other zones
What cost implications can be involved with using Availability zones?
- There could be a cost to duplicating services and transferring data between zones
What types of resource are availability zones typically used for
- Availability Zones are primarily for VMs, managed disks, load balancing and SQL DBs
What categories of services that fall into availability zones are there
Zonal Services
Zone Redundant Services
What is meant by a Zonal Service within Availability Zones
- Zonal Services - Pin resource to specific zone (i.e. VMs, managed disks, IPs)
What is meant by Zone Redundant Services within Availability Zones
- Zone Redundant Services - Platform replicates automatically across zones (i.e. zone redundant storage and SQL dbs)
Availability zones are created using one or more data centers, there are a minimum of 3 zones in a region. However a large enough disaster could cause an outage to effect even two data centers. What Azure solution could be used to mitigate this risk
Region Pairs
What is a region pair?
- Each Azure region is paired with another within the same geography (US, Europe, Asia etc)
What is the minimum distance that a region pair must be from the other
At least 300 miles
Region Pairs in Azure allows replication of resources (Such as VM Storage) across a geography, this is to minimize interruptions that may occur due to what…
- Allowing replication of resources (such as VM storage) across a geography minimizes interruptions that may occur due to natural disasters, civil unrest, power or network outage affecting both regions at once.
If one region pair was affected by a natural disaster, what would happen
If a region pair was affected by a natural disaster, services would failover automatically to the other region pair
What are some of the additional advantages of region pairs
- In extensive Azure outage - 1x region of every region pair is prioritized for restore
- Planned Azure updates are rolled out to paired regions one at a time to minimize downtime and risk.
- Data resides in same geography as pair (except Brazil South) for tax and law enforcement jurisdiction purposes
- A broad set of distributed DC’s allows Azure to provide a high guarantee of availability.
What do SLAs capture
SLAs capture terms that define standards that apply to azure.
There are SLAs to for specific Azure products and services.
SLAs describe MS commitment to providing Azure customers with what
Specific performance standards
- SLAs specify what happens if product or service fails to perform to the SLAs specification. However what sort of SLAs are provided for most Free/Shared Tiers
None
What 3 Characteristics are there for SLAs for Azure Products and Services
Performance Targets
Uptime and Connectivity Gurantees
Service Credits
What is meant by SLA Performance Targets
- SLA defines performance targets for product or service these are specific to each azure product of services (e.g. uptime guarantees of connectivity rates)
What is meant by SLA Uptime and Connectivity Guarantees
- typical SLA commits to between 99.9% or 99.999% (i.e. 3 OR 5 nines)
- Example - Cosmos DB has SLA of 99.999% uptime with low latency commitments of 10 Miliseconds on DB Reads and Writes
What is meant by Service credits
SLAs also describe the response if an Azure Product or Service fails to perform to specification. Customer may have a discount applied to there bill as compensation.
Define a composite SLA and what it can result in
Combining SLAs across different service offerings results in a composite SLA.
This can result in a higher or lower uptime value depending on your application architecture
How can you improve a composite SLA
You can improve composite SLA by creating independent failback paths
You can create your own SLAs to meet your own performance targets to suit you specific Azure application. What is this known as
Application SLA
What do you need to know to build and efficient, reliable solution
- Building an efficient, reliable solution requires knowing your workload requirements and selecting/provisioning azure products/services according to those requirements.
What existing data in Azure will help create an achievable Application SLA
- Azure SLAs define performance targets for Azure products/services within your solution. This will help create achievable application SLAs
What is meant by the term resiliency
Resiliency - ability to recover from failure and continue to function
- Not avoiding failures but responding to them in a away that avoids downtime or data loss.
What is the goal of resiliency
The goal of resiliency is to return to a fully functioning state following a failure
What are two key components of resiliency
HA (High availability) and DR (Disaster Recovery) are 2 key components of resiliencey.
When designing architecture and designing for resiliency what should you perform, and what is it’s goal
- When designing architecture - design for resiliency, your should perform a FMA (Failure Mode Analysis) which goal is to identify possible points of failure and define how to respond to such failures.
What does availability refer to (try not to use the word available)
- Availability refers to time system is functional and working, maximizing availability requires implementing measures to prevent possible service failures.
Increasing availability is expensive and increases what…
Complexity
As complexity grows with higher availability solutions more services depend on each other this makes it easy to overlook….
- As complexity grows more services depend on each other, making it easier to overlook possible failure points
An SLA of 99.999% allows for approx how much down time a year
- SLA of 99.999% allows for approx 5mins of downtime a year.
For 99.99% targets (4-nines) manually recovering from failures may not be enough, what might you need to investigate instead
Investigate Self diganosing and Self healing solutions
Why should you consider the time window against which your SLA targets are measured
Consider time window against which SLA targets are measured, a smaller window will mean tighter tolerances, hourly or daily uptime measurements might not allow for achievable targets.
