Kubernetes Deployment and Resources (Quiz 3)
Study these flashcardsFree · spaced repetition · iPhone, Android, Mac & web
All cards
1
What is a deployment strategy
describes how we replace or introduce a new version of an application without uneccesarily disrupting users
2
Q: What is the key benefit of a shadow deployment?
Tests a new application version using real production traffic without impacting users, because the new version's responses are discarded. This allows teams to detect bugs and performance issues before release.
3
True or False) With serverless, a function can scale down to zero
instances when there is no traffic, so you pay only for the time your code
actually runs
instances when there is no traffic, so you pay only for the time your code
actually runs
Serverless platforms scale to zero when idle and scale out
automatically under load. Billing is based on actual execution time and the
number of times the function runs, not on idle capacity
automatically under load. Billing is based on actual execution time and the
number of times the function runs, not on idle capacity
4
(True or False) In serverless computing, the developer is responsible for
provisioning, patching and managing the servers that run the functions
provisioning, patching and managing the servers that run the functions
False The cloud provider provisions, patches, scales and manages
the servers. The developer only writes and deploys the function code.
the servers. The developer only writes and deploys the function code.
5
How do you know something is canary
t: Gradually increases the percentage of users receiving the new version while monitoring performance and limiting the impact of potential bugs.
6
What is a pod
The smallest deployable unit in Kubernetes. It runs one or more containers.
7
What is a Replica
Another instance of a Pod running the application
8
What is a DEployment
A Kubernetes resource that declares the desired application version and number of replicas.
9
What is a ReplicaSet
A resource that maintains the desired number of matching Pods.
10
What is a Service
A stable network access point that directs traffic to matching Pods.
11
What is Ingress
A Kubernetes resource containing rules for routing external HTTP/HTTPS requests to Services
12
What does the Deployment control the Pods through?
A repliceset
13
What provides network access to eligible pods
Service
14
What ensures the exact number of identical pods are alive at any given moment?
ReplicaSet
15
What is a rollback?
Returning an application to a previous working version after an unsuccessful update
16
What is a readiness probe?
A kubernetes check that determines whether a container is ready to receive traffic
17
Why does a team choose a rolling update?
Update continuously while some Old Pods remain available
18
What are the key benefits of a rolling update?
High availability, manageable resource usage
19
Why does a team choose Canary deployment?
To see how V2 behaves with a small amount of live traffic
20
What is the key benfit of Canary?
Limits the number of users affected by bugs
21
Why does a team choose a Blue/Green Deployment?
Prepare an entire replacement environment before switching traffic?
22
Note the three step process of bluereen deployment
Install and test new software on Greenenvironment while Blue keeps running normally for users
Once green passes all your checks, update load balancer to send all incoming user traffic to Green instead of Blue
If something breaks in Green you instantly flip the load balancer back to blue
Once green passes all your checks, update load balancer to send all incoming user traffic to Green instead of Blue
If something breaks in Green you instantly flip the load balancer back to blue
23
What does the Blue Deployment act as in a Blue/Green Deployment ?
An intermediate safety net
24
What are the key benefits of bluereen deployment
Near zero downtime releases paired with instant, risk free rollbacks
25
Why does a team choose A/B testing deployment
Compare two versions of an application, feature, or asset side by side in a live environment to determine which performs better
26
What is the key benefit of A/B testing deployment? Describe it...
Data Drive decision making which replaces internal bias and guesswork with objective user behavior
27
Describe Shadow deployment -> steps
When a request arrives, the system duplicates it and sends both versions
The live version processes the request and returns the actual response to the user
The shadow version processes the same request in the background, logs its output and discards the response
The live version processes the request and returns the actual response to the user
The shadow version processes the same request in the background, logs its output and discards the response
28
What is the default strategy when you change a kubernetes deployment
Rolling
29
Describe 4 stages of Rolling Update
Three V1 pods serving users
Create a V2 pod while V1 remains available
Wait until V2 is available before sending service traffic
Kubernetes reduces old ReplicaSet and continue until V2 replaces V1
Create a V2 pod while V1 remains available
Wait until V2 is available before sending service traffic
Kubernetes reduces old ReplicaSet and continue until V2 replaces V1
30
What does Service give ?
Gives the application a stable internal network identity even though Individual Pods could change
31
What's the difference between Ingress and Service ?
Ingress decides which Service an external request should reach while service directs the request o an eligible Pod
32
What layer of the OSI model does a Kubernetes Ingress operate at, and what is its primary function?
It operates at Layer 7 and it acts as a reverse proxy and smart router sitting at the edge of the network to securely route external HTTP/S traffic to internal cluster based on URL paths or host names
33
What is runtime?
The environment needed to execute code, such as a Python runtime.
34
What is a serverless function?
A small piece of code that runs when triggered by an event
35
What do cloud providers do in serverless?
manage underlying infrastructure and automatically scale function instances based on demand
36
What is a common miconception regarding serverless functions?
Serverless doesn't mean no servers exist it means you don't manage them
37
A Kubernetes application has three V1 Pods. When V2 is deployed, Kubernetes launches a new Pod, confirms it is ready, and removes an old Pod. What strategy is this, and why?
A rolling update gradually replaces old Pods with new ones. Kubernetes checks that new Pods are ready before removing old Pods, maintaining application availability and allowing rollback if needed
38
An application has 10 replicas of V1 and 2 replicas of V2, both eligible to receive user traffic. Why might a team deliberately configure it this way?
This is a canary release. By running fewer V2 replicas than V1 replicas, the team exposes a smaller proportion of users to V2. They can monitor its performance, increase its traffic if stable, or roll back if errors occur.
39
A user visits example.com/checkout. Explain the responsibilities of Ingress, Service, and Pod in getting that request handled.
Ingress examines the requested URL path and routes the request to the appropriate Service. The Service provides a stable network endpoint and forwards traffic to an eligible Pod. The Pod runs the application code and processes the request
40
Which Kubernetes mechanism, as described in the text, ensures a container is prepared to handle incoming requests before it's added to a service's traffic pool?
Readiness Probe — ensures the container is ready before receiving traffic.
41
What is a primary advantage of using ReplicaSets during a rolling update process?
Simplified rollback. ReplicaSets retain the configuration needed to recreate Pods from previous versions.
42
When a rolling update is triggered, how does Kubernetes manage resource consumption during the update process?
In the textbook example, it starts one new replica, verifies readiness, then terminates an old replica. This avoids a large resource surge.
43
What is the main purpose of a canary release?
Test a new version with a small subset of real production traffic before full rollout.
44
Which common method allows you to manage the proportion of traffic directed to different service versions during a Kubernetes canary release?
Adjusting the replica count for each version's Deployment.
45
When are bluereen deployments most suitable?
When clients cannot safely consume the old and new versions simultaneously.
46
How is traffic switched from the blue environment to the green environment in a bluereen deployment using Kubernetes?
Update the Service's selector labels to target the new environment's Pods.
47
What important consideration should be taken into account regarding stateful applications when performing bluereen deployments?
State data must be migrated or made accessible to the new environment through persistent storage.
48
What is the primary goal of A/B testing in the context of service deployments?
Compare two versions simultaneously to determine which performs better using user feedback.
49
How can you route traffic to different versions of a service for A/B testing in Kubernetes, according to the text?
Define an Ingress resource with multiple rules, each directing traffic to a different Service.
50
What is the primary function of Kubernetes Ingress?
Provide routing rules for external HTTP/HTTPS traffic accessing Services inside a Kubernetes cluster.
51
True or False) As the illustration shows, serverless moves virtualization up to the runtime level, making the individual function, rather than the VM or the application, the unit of scaling.
True — hypervisors scale VMs, containers scale applications, and FaaS scales individual functions.
52
Define a rolling update and identify three benefits.
Gradually replaces old pods with new ones
Benefits: High availability, controlled resource consumption, and simplified rollback.
Benefits: High availability, controlled resource consumption, and simplified rollback.
53
What's the fundamental difference between a canary and a bluereen deployment?
Canary: Gradually exposes V2 to increasing percentages of live traffic.
Blue/Green: Switches production traffic entirely between two environments.
Blue/Green: Switches production traffic entirely between two environments.
54
What is a shadow deployment, and how does it differ from a canary release?
V2 receives copies of real requests, but its responses are discarded.
Unlike canary, users never receive V2's responses during testing.
Unlike canary, users never receive V2's responses during testing.
55
Define Function as a Service (FaaS) and explain how scaling works.
Event-driven code executed on provider-managed infrastructure.
Functions automatically scale with demand, potentially down to zero when idle.
Functions automatically scale with demand, potentially down to zero when idle.
56
Define abstraction and provide a Kubernetes example.
Hiding underlying implementation details behind a simpler interface.
Example: A Kubernetes Service provides a stable endpoint while hiding changing Pod IPs.
Example: A Kubernetes Service provides a stable endpoint while hiding changing Pod IPs.
57
What does service structure mean in cloud-native application development?
How an application is divided into services with defined responsibilities and communication.
58
Suppose a Deployment has three V2 Pods running, and its previous V1 ReplicaSet has zero Pods. Why can Kubernetes still restore V1 without you manually recreating its configuration?
Because V1's replica set still knows how to recereate the pods how they were originally desired
59
What is a service selector
A configuration field used to map traffic from a Service to a specific group of pods
60
Why might a devops team be hesitant to deploy a new version of their service?
a) Concerns about hardware compatibility.
b) Risk and potential disruptions from changes.
c) Lack of interest in innovation.
d) Mandatory downtime policies
a) Concerns about hardware compatibility.
b) Risk and potential disruptions from changes.
c) Lack of interest in innovation.
d) Mandatory downtime policies
B
61
What most directly translates to increased business value in software development?
a) Using the latest technology stack.
b) Frequent releases with bug fixes and new features.
c) Having a highly skilled development team.
d) Reducing infrastructure co
a) Using the latest technology stack.
b) Frequent releases with bug fixes and new features.
c) Having a highly skilled development team.
d) Reducing infrastructure co
B
62
Which of these is a fundamental Kubernetes resource used to expose applications?
a) Pod
b) Namespace
c) Service
d) ConfigMap
a) Pod
b) Namespace
c) Service
d) ConfigMap
C
63
Which Kubernetes mechanism, as described in the text, ensures a container is
prepared to handle incoming requests before it's added to a service's traffic pool?
a) Liveness Probe
b) Readiness Probe
c) Startup Probe
d) Health Check
prepared to handle incoming requests before it's added to a service's traffic pool?
a) Liveness Probe
b) Readiness Probe
c) Startup Probe
d) Health Check
B
64
9. What is a primary advantage of using ReplicaSets during a rolling update process?
a) Automatic pod scaling based on traffic demands.
b) Simplified rollback to previous deployment configurations.
c) Provision of stable IP addresses for pods.
d) Automated network infrastructure configuration.
a) Automatic pod scaling based on traffic demands.
b) Simplified rollback to previous deployment configurations.
c) Provision of stable IP addresses for pods.
d) Automated network infrastructure configuration.
B
65
When a rolling update is triggered, how does Kubernetes manage resource consumption during the update process?
a) It immediately starts all new replicas, causing a surge in resource usage.
b) It starts a single new replica, validates its readiness, and then terminates an old replica.
c) It pauses all replicas until the update is complete, minimizing resource consumption.
d) It doubles the resources allocated to the deployment to ensure a smooth transition.
a) It immediately starts all new replicas, causing a surge in resource usage.
b) It starts a single new replica, validates its readiness, and then terminates an old replica.
c) It pauses all replicas until the update is complete, minimizing resource consumption.
d) It doubles the resources allocated to the deployment to ensure a smooth transition.
B
66
What is the main purpose of a canary release?
a) To test a new version of a service with a subset of live traffic before a full rollout.
b) To automatically deploy new versions to all users as quickly as possible.
c) To ensure zero downtime during deployments.
d) To reduce resource consumption during updates.
a) To test a new version of a service with a subset of live traffic before a full rollout.
b) To automatically deploy new versions to all users as quickly as possible.
c) To ensure zero downtime during deployments.
d) To reduce resource consumption during updates.
A
67
Which common method allows you to manage the proportion of traffic directed to different service versions during a Kubernetes canary release?
a) Implementing a dedicated traffic management system.
b) Defining intricate Ingress routing rules.
c) Manually assigning traffic to individual pods.
d) Modifying the replica count for each deployment version
a) Implementing a dedicated traffic management system.
b) Defining intricate Ingress routing rules.
c) Manually assigning traffic to individual pods.
d) Modifying the replica count for each deployment version
D
68
When are bluereen deployments most suitable?
a) When minimizing resource consumption during deployments.
b) When upgrading from one version to the next, and users/clients cannot consume both versions simultaneously.
c) When different versions can be tested simultaneously.
d) When there are no breaking changes between versions.
a) When minimizing resource consumption during deployments.
b) When upgrading from one version to the next, and users/clients cannot consume both versions simultaneously.
c) When different versions can be tested simultaneously.
d) When there are no breaking changes between versions.
B
69
How is traffic switched from the blue environment to the green environment in a bluereen deployment using Kubernetes?
a) By gradually shifting traffic using weighted routing rules in the Ingress.
b) By creating a new service resource that points to the green environment.
c) By updating the selector labels in the service resource to match the new version's labels.
d) By manually redirecting traffic using a load balancer.
a) By gradually shifting traffic using weighted routing rules in the Ingress.
b) By creating a new service resource that points to the green environment.
c) By updating the selector labels in the service resource to match the new version's labels.
d) By manually redirecting traffic using a load balancer.
C
70
What important consideration should be taken into account regarding stateful applications when performing bluereen deployments?
a) State data should be migrated from the blue environment to persistent storage so the green environment can access it.
b) The service must be able to automatically roll back to the previous version in case of errors.
c) Stateful applications are not compatible with bluereen deployments.
d) The state data will automatically be replicated from the blue environment to the green environment.
a) State data should be migrated from the blue environment to persistent storage so the green environment can access it.
b) The service must be able to automatically roll back to the previous version in case of errors.
c) Stateful applications are not compatible with bluereen deployments.
d) The state data will automatically be replicated from the blue environment to the green environment.
A
71
What is the primary goal of A/B testing in the context of service deployments?
a) To automatically roll back to a previous version if errors occur.
b) To reduce resource consumption by running only one version of the service at a time.
c) To ensure zero downtime during updates.
d) To test different versions of a service simultaneously to determine which performs better based on user feedback
a) To automatically roll back to a previous version if errors occur.
b) To reduce resource consumption by running only one version of the service at a time.
c) To ensure zero downtime during updates.
d) To test different versions of a service simultaneously to determine which performs better based on user feedback
D
72
How can you route traffic to different versions of a service for A/B testing in Kubernetes, according to the text?
a) By using a single service and dynamically changing the target pods.
b) By defining an Ingress resource with multiple rules, each pointing to a different service.
c) By adjusting the number of replicas for each deployment.
d) By manually directing traffic to specific pods based on user IDs
a) By using a single service and dynamically changing the target pods.
b) By defining an Ingress resource with multiple rules, each pointing to a different service.
c) By adjusting the number of replicas for each deployment.
d) By manually directing traffic to specific pods based on user IDs
B
73
What is the primary function of Kubernetes Ingress?
a) To automatically scale the number of pods in a deployment.
b) To monitor the health and performance of applications running in Kubernetes.
c) To provide routing rules for external access to services in a Kubernetes cluster.
d) To manage internal network policies within the cluster.
a) To automatically scale the number of pods in a deployment.
b) To monitor the health and performance of applications running in Kubernetes.
c) To provide routing rules for external access to services in a Kubernetes cluster.
d) To manage internal network policies within the cluster.
C
74
What is often cited as an advantage of using Kubernetes Ingress over directly employing a cloud provider's Load Balancer, particularly in terms of management and cost?
a) Ingress centralizes traffic routing rules within the Kubernetes cluster, potentially simplifying management and reducing expenses compared to managing an external Load Balancer separately.
b) Cloud Load Balancers lack the capacity for SSL termination.
c) Cloud Load Balancers are strictly limited to non-production scenarios.
d) Ingress inherently offers superior performance and reduced latency.
a) Ingress centralizes traffic routing rules within the Kubernetes cluster, potentially simplifying management and reducing expenses compared to managing an external Load Balancer separately.
b) Cloud Load Balancers lack the capacity for SSL termination.
c) Cloud Load Balancers are strictly limited to non-production scenarios.
d) Ingress inherently offers superior performance and reduced latency.
A
75
Which of the following is a benefit of using Ingress in a Kubernetes cluster?
a) It automatically manages the underlying hardware resources.
b) It provides a dedicated IP address for each Pod in the cluster.
c) It automatically creates backups of application data.
d) It consolidates routing rules into a single resource, simplifying configuration and management.
a) It automatically manages the underlying hardware resources.
b) It provides a dedicated IP address for each Pod in the cluster.
c) It automatically creates backups of application data.
d) It consolidates routing rules into a single resource, simplifying configuration and management.
D
76
Based on the Function as a Service illustration above, which statement best describes how a FaaS platform behaves as the number of events changes?
a) The developer must provision the servers, operating system and runtime before the platform can respond to events.
b) When no events arrive, the function scales to zero and costs nothing. When a spike of events arrives, the platform automatically runs many copies of the function in parallel, and you pay only for the actual run time.
c) A fixed number of virtual machines stays running at all times, and the developer must add more VMs manually when a spike of events occurs.
d) The function runs continuously in a single container, and every event waits in line to be processed one at a time.
a) The developer must provision the servers, operating system and runtime before the platform can respond to events.
b) When no events arrive, the function scales to zero and costs nothing. When a spike of events arrives, the platform automatically runs many copies of the function in parallel, and you pay only for the actual run time.
c) A fixed number of virtual machines stays running at all times, and the developer must add more VMs manually when a spike of events occurs.
d) The function runs continuously in a single container, and every event waits in line to be processed one at a time.
B
More Computer science flashcards
Internet & WWW Study GuideComputer science · EnglishSgLinux 2344 FinalComputer science · EnglishCarterImpact of Computing Study GuideComputer science · EnglishSgCS 325 T7 L4,5,6Computer science · EnglishClement1.3.4 WEB TECHNOLOGIESComputer science · EnglishSabs2023Lecture 1 YuComputer science · EnglishDiandreiLinux Commands for Comptia A+ 1102Computer science · Englishan OutLearn studentc programmingComputer science · EnglishHaseeb1968Inheritance, Composition, & Data EfficComputer science · EnglishNijelousDirectory Services and AttributesComputer science · EnglishAydenInformation Technology BasicsComputer science · EnglishSamIT Hardware and Mobile TroubleshootingComputer science · EnglishRichard
Study these decks in the app
Free on iPhone, Android, Mac and the web. Add any deck to your account in one tap.
