Monitoring & Reporting Flashcards by Andrews Kuhn Silva

What is cloud watch?

For monitor resources on AWS and Apps running on EC2

How well did you know this?

Not at all

Perfectly

What Cloud Watch can monitor?

Compute Resources, Storage And Content delivery, Database and Anlaytics and Others

How well did you know this?

Not at all

Perfectly

What types of compute resources that Cloud Watch can monitor ?

Scaling Groups, ELB and Route53 Health Checks.

How well did you know this?

Not at all

Perfectly

What types of Storage and content delivery that Cloud Watch can monitor ?

EBS, Storage Gateways and CloudFront

How well did you know this?

Not at all

Perfectly

What types of Database and Analytics that Cloud Watch can monitor ?

Dynamodb, elastic cache, RDS, Elastic mapReduce job flows, RedShift

How well did you know this?

Not at all

Perfectly

What types of others that Cloud Watch can monitor ?

Sns topics, sqs queue, OpsWorks, CloudWatch logs, AWS Bills

How well did you know this?

Not at all

Perfectly

What types of metrics that cloud watch can monitor in EC2 ?

# Host level
  - CPU
  - Network
  - Disk I/O
  - Status Check
# Custom metrics:
  - Ram utilization
  - How much is left of space from Disk

How well did you know this?

Not at all

Perfectly

What is the min granularity that custom metrics can use?

1 minute

How well did you know this?

Not at all

Perfectly

What is the default interval for monitor Host level EC2?

5 minutes but can be 1 minute

How well did you know this?

Not at all

Perfectly

What are the types that states check can be applied?

EC2 Host and EC2 instance

How well did you know this?

Not at all

Perfectly

What is the default time that cloud watch metrics are stored?

Indefinitely, but can be customized to expire in 2 weeks, 2 month and etc…

How well did you know this?

Not at all

Perfectly

What is the state that EC2 and ELB can recover logs?

Both EC2 and ELB Logs can be recovered form terminated state.

How well did you know this?

Not at all

Perfectly

What is On premise Cloud Watch?

Is when cloud watch is used on local DataCenter/Server with SSM Agent or CloudWatch agent.

How well did you know this?

Not at all

Perfectly

What are the types of EBS and uses?

#SSD(GP2)
  - Most used
  - Can Boot System
#Provisioned IOP SSD(io1)
  - Great Volume of I/O
  - Used for > 16,000 IOP 
#HDD(st1)
  - For Cheap disk like Big Data, logs, Data warehouse
  - Cannot use for boot
#Cold HDD(SC1)
  -  Super low cost
  - Infrequent acess
  - Cannot be used for boot

How well did you know this?

Not at all

Perfectly

How GP2 credits works?

Gain 3 IOPs per GB and if is not > 3000 IOPS can use for 30 minutes burst of the left 3000 IOPS, and restore every minute not used.

How well did you know this?

Not at all

Perfectly

What is pre-warming on EBS?

Is no longer necessary, but if restore ebs from snapshot is need a pre-warming, so every block of the disk need to be readed.

How well did you know this?

Not at all

Perfectly

What are the metrics on EBS that can be used and for what?

# Read/Write IOPS, If > 16,000/5.6 TB switch to io1/provisioned iops SSD
# Volume Queue length, best case in  0,  is queue of operations of I/O, must increase the IOPS

What are the volume check status and the problems come for this?

OK, Warning, impairede, insufficient data.

Degraded/Severely degraded === Warning
Stalled / Not Available === Impaired

How modify the size of EBS?

Its can be modified the size and IOPS for io1, but on change size must ssh on instance and extend the Volume.

What are the three types of ELB?

# Aplication load balancer - For general use, like http and https.
# Network load balancer - High I/O for TCP
# Classic load balancer - Its like Application, but previous generation

What are the types of metrics for ELB?

#Cloud Watch metric - how many 500, latency etc.
#Acess logs - Who acess the resource (IP), what is the path, request and response.
#Request tracing -  Is for aplication load balancer, add Header for trace the origin.
#CloudTrail logs -

What is the difference between CloudWatch, CloudTrail and Config?

#CW - Monitor the performance, response time, cpu utilization, ram utilization. SPORT perfomance
#CT - API calls to AWS Resources, ip and who(audiction). SPY 
#C - Record the aws Enviroment resouces and notify on changes and break rules. CAMERAS AND CCTV

What are the types of metrics on memory cache?

Cpu utilization, Swap usage, Evictions and concurrent connections.

What are cautions need to monitor cpu utilization on memory cache?

memcache

Multithread
Handle loads of 90%, if > 90% add more nodes
#Redis
Not multithread
90/Cores to see if needs to scale, 90 / 4 cores -> 22,5% usage to scale

What are cautions need to monitor swap usage on memory cache?

``` #memcache - +/- 0 of swap and not exceed 50mb, if > 50mb raise memcache_conections_overhead #Redis - no swap metrics ```

What are cautions need to evictions usage on memory cache?

#memcache - Can scale up the machine(more memory) or scale out more nodes #Redis - Only scale out

What are cautions need to concurrent connections usage on memory cache?

If had spike on this, need to see if was spike on traffic or the app is not droping connections

What you need to know to make cloudwatch dashboards?

Dashboards are global, but if you need get a specific resource data you need change the region that the resource are in.

What are the types of billing alarms

``` # Cost - configure a forecast cost and trigger alarm after threasold percentage of the cost passed # Use - Configure a forecast cost and trigger alarm after threasold percentage of the cost passed for group or resource. # Reservation - For reservation usage. ```

What you can do with AWS organizations

``` #Centralize Policies of allow and deny resources between accounts #Acess control to AWS resouces #Automate Aws accounts creation and managent in groups #Consolidate billing ```

What are the types of Tag managers and the diferences?

``` #Classic resources groups - Can view data about groups tagged with the same tag, can be global or region #AWS system manager - Can view insight and do automation on tag groups but is regional. ```

What are the model price for EC2?

``` #OnDemand - Pay as you go hourly or second #Reserved - Reserve a capacity of ec2 offer significant discount, term of 1 year or 3 years #Spot - Bid price for instances, recommended for apps that have start and finish flexible #Dedicated - EC2 dedicated for your, not sharing. Can have discount for uses that has software licenses (server-bound) ```

What are AWS Config

Is for monitor changes and rules on resources on AWS region based, can be trigger periodic or by changes, have custom and pre-created RULES to apply to the resources.

What are the permissions needed for AWS config?

IAM role with read only permission to resources monitored, write permission for s3 bucket for log and publish to SNS

Who can access AWS config?

Only Sys admins with full access, and Aws config need only read access for monitored resources.

What CloudTrail can do with Aws Config?

You can setup cloudTrail to log, changes on AWS config, like disabling record/rule of some resource

Tell one example of Config, CloudTrail, CloudWatch.

``` #Config - See the state of security group 3 weeks before #CloudTrail - See who change the security group 3 weeks before #CloudWatch - See the CPU and Memory utilization ```

What are the types of Health Dashboard and usage?

``` #Service Health Dashboard - Is continental, is health of AWS services. #Personal Health Dashboard - Is global, are issues that can affect services that you are using on AWS. ```

What are the types of metric granularity and example?

``` # 1 min is detailed monitoring - If asks for 3 minutes monitor interval # 5 min is standard monitor - if asks for 10 minutes monitor interval ```

Your application got a lot of 500 two week before, How can you debug this?

Just use Acess logs on the ELB stored on s3 bucket, the deleted instances logs are there.

What is IOPS?

Input/Output Operations per second