Deployment and Provisioning Flashcards

1
Q

Common reasons why EC2 instnaces may fail to launch

  • InstanceLimitExceeded error
  • you have reached the limit on teh number of instnaces you can launch in a region
  • aws sets default limits on the number of instances you can run on a per-region basis - 20 by default
A

t

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

InsufficientInstanceCapacity error

-AWS does not currently have enough available On-Demand capacity to service your request

A

y

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

____ allows you to create storage volumes and attach them to your EC2 instances.

A

EBS

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

___ can be used to craete.a file system, run a database, run an operation system, etc.

A

EBS

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

____ backend storage - used fr OS and databases which can be IO intensive.

A

SSD

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

2 different variants of SSD:

gp2 ___ - boot volumes

io1 ___ - IO intensive, NoSQL/relational databases, latency sensitive workloads

____ used to benchmark performance for SSD volumes

A

general purpose

Provisioned IOPS

IOPS

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

IOPS capability is dependent on the size of your volume:

gp2 volumes: (minimum 100 IOPS) 3 IOPS/GB up to a max of _____ IOPS

io1 volumes: 50 IOPS/GB to a amx of _____ IOPS

A

16,000

64,000

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

what happens if you are using gp2 and your workload exceeds the IOPS limit of teh gp2 volume you have provisioned?

A

you will start to get your IO requests queing

depending on your apps seneitivity to IOPS and latency, you may see your app becoming slow.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Hitting the IOPS limit of our gp2 Volume

what can you do about it?

explain the 2 approaches to address hitting the IOPS limit:

A
  • For gp2, you can increase the size of your volume - but if your volume is already 5.2T or moer, you will have already reached teh 16,000 IOPS limit for gp2 volumes
  • If you need more tahn 16,000 IOPS, you will need to change your storage class to Provisioned IOPS
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

A bastion host is located in your ____ subnet

A

public

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

a bastion host allows you to connect your ec2 instances using ___ or ___ protocols

A

ssh or rdp

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

T or F

you can log in to the Bastion host over the internet, from your desktop

A

T

If the settings allow it

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

T or F

the bastion host is used to initiate an SSH/RDP session over the private subnet to your ec2 instances in teh private subnet

A

T

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

____ load balancers scale automatically to adapt to your workload

A

application

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

xx messages indicate that something has gone wrong on the client side

A

400

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

xx messages relate to server side errors

A

500

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

= Bad/malformed request - ie: the header is malformed

A

400

18
Q

Unauthorized - user access denied

A

401

19
Q
  • forbidden - request bloked by WAF access control list
A

403

20
Q
  • client closed connection before the load balancer could repond, client timeout period may be too short
A

406

21
Q
  • load balancer received an X-Forwarded-For request headerwith >30 IP addresses - similar to a malformed request
A

463

22
Q
  • Internal srver error iee: error on the load balander
A

500

23
Q
  • bad gateway - ie: an app server closed the connection or sent back a malformed response
A

502

24
Q
  • server unavailble - no registered targets
A

503

25
Q
  • gateway timeout - ie: ap is not responding - problem with your web server, app server or database
A

504

26
Q
  • Unauthorized - received an error code from the ID provider when trying to authenticate a user
A

561

27
Q

____ load balancers publish metrics to CW for the load balancer and for the backend instances

A

elastic

28
Q

t or F

ELB CW metrics help to verify taht your system is performing as expected

A

T

29
Q

ELB CW metrics - overall health

_____ number of unsuccessfl conenctions to backend instances

_____ number of healthy insatnces registered

___ number of unhealthy instances

____2xx,3xx,4xx,5xx

A

BackendConnectionErrors

HealthyHostCount

UnHealthyHostCount

HTTPCode_Backend

30
Q

____ number of seconds taken for ergistered instance to respond/connect

A

Latency

31
Q

____ number of requests completed/connections made during teh specified interval (1 or 5 min)

A

RequestCount

32
Q

____ number of pending requests, max queue size is 2014, additional requests will be rejected (Classic only)

A

SurgeQueueLength

33
Q

_____ number of requests being rejected because the surge queue is full (Classic only)

A

SpilloverCount

34
Q

____ is a mangment tool which gives you visibility and control over your AWS infrastracture.

It integrateswith CW allowing you to view your dashboards, view operational data and detect problems.

Includes ___ which automates operational tasks across resources - ie: security patching, package installs.

Organize your inventory, grouping resources together by applicaiton or environment - including on prem systems.

A

Systems Manager (SSM)

Run Command

35
Q

_______

allows you to run pre-defined commands on one or more EC2 instances.

  • stop, restart, terminate, re-size instance
  • attach/detach EBS volumes
  • create snapshots, abckup dynamoDB tables
  • apply patches and upadtes
  • run an ansible playbook
  • run a shell script
A

run command

36
Q

3 types of placement groups:

___ instances are all created in a single AZ

___ instances are created in logical segments called partitions, each located in a separate rack(s), with independent network and power.

____ each instance is created in a separate rack, with independent network and power.

A

CLuster, Partition, Spread

37
Q

___ placement groups

  • all instances created in same AZ
  • all nodes within the placement group can talk to all other nodes within the placement group at the full line rate of 10Gbps
  • Not great availability (ideally, you’d want a copy of your stack in another AZ as well)
  • great if the separate compnents in your app require high network throughput
A

cluster

38
Q

_____ placement group

  • instances are created in logical segments called partitions, can be multi-AZ
  • each partition is in a separate rack(s), with independent network and power
  • partitions cannot share racks which helps reduce the impact of a failure
  • great for large distribuetd/replicated workloads, such as HDFS< HBase, and Cassandra
A

Partition

39
Q

___ placement groups

  • instances placed on distinc racks, each with its own network and power source
  • recommend for apps with small number of critical intances taht should be separated
  • reduces teh risk of simultaneous failures when instances share the same racks.
  • gurantees your instances are places in different racks with isolated power and networking.
A

spread

40
Q
A