GCP exam questions, as far as I remember

• General:
  Total questions : 50
  Two case studies : 20 questions
  CI/CD - 1
  Storage/Database - 15
  General google services - 3
  Appengine vs GKE vs GCE usage -
  Preemptible instances / AutoScaling  / Global network
  Monitoring / stackdriver  - 5
  
  
  Exam subjects:
  CI/CD:
  • The correct format of the template for cloud deployment  - yaml (+), jinja (+ )
  template for cloud deployment
  Storage:
  • Difference between storage tiers
  • Quite a bit questions that ask about what is the best storage / database to use for customer needs.
  • VM Disks difference, which is faster, which is more reliable
  
  
  GCE:
  • AutoScaling triggers
  • Connection between 2 GCP projects - sharing resources, vpn connections, vpc
  • Instance Tags for firewall
  • ELB question about difference between HTTP(S) and UDP TCP (one is regional and one is cross regional)
  • Preemptable instaces (VMs)  question - usage
  • Company wants to run their J2EE application, where is the best to deploy it?
  • GCE
  • GAE Standard / Flex
  • GKE

  • External IP addresses - reserving

  • Slow connection between on-prem and GCP - direct connect, multivpn,....
  • Network : global and spans different zones (regions contain zones)
  • Default firewall - allows alot. Custom firewall - everything is forbidden
  • Company X want move to cloud and they have X configurations on-prem. What is the best way to build their environment in the cloud?
  • Start with smallest VMs and scale up
  • Start with biggest VMs and scale down
  • Go to their on-prem, see their configuration and build similar in the cloud using custom VM conf
  • Estimate VMs (cpu, ram…) they will need and build costume VMs according to estimation
  • Tags
  
  
  Monitoring/Logging
  • Developers are not satisfied with logging of their application. What would you suggest to them?
  • Company wants monitor and debug. What stackdriver options will they need? And than you are presented with some real, not real options and some are real but not needed.
  
  
  DBs / BigData
  • Quite a bit questions that ask about what is the best storage / database to use for customer needs.
  • Company can’t tolerate ANY downtime at all. How do you copy their database
  from on-prem to google cloud?
  • Cloud SQL - Spanner - BigQuery ( understand the usage, according to the size of the project, company and amount of data)
  • Cloud SQL doesn’t work with private IP, only public (currently)
  • Developers know SQL but you need to operate with noSQL DB. What you use?
  • If you see Spark or / and Hadoop - the answer is Dataproc
  • 1-2 questions about Dataflow
  • IoT - weather station data streaming - Pub/sub or Dataflow
  • ColdLine, nearline, difference in prices, frequency and cost efficiency
  
  
  
  Security:
  • IAM roles and permissions difference
  • Company has DEV and Prod env. Split them into different projects, not accounts
  • Can you create account for GCP for a user, who has no gmail or g suite account
  • G suite and GCP connection - IAM usage
  
  
  
  
  Other services / General cloud knowledge:
  • Cloud DNS - the correct name and what it used for
  • Lables
  • People are complaining that news feed they see got all news marked as read.
    * INSERT PYTHON CODE*
    What could cause it?
  
  
  
  
  
  GKE & APP ENGINE - Coursera , course #4
  
  
  GKE
  • Master node…….. One and autom. Created in GKE
  • Full resource consumption - what to use to use 100% of resources  (containers)
  
  
  APP ENGINE:
  • Flex and standard env in GAE : difference in usage
  • Ability to run containers (GAE flex)
  • Opsless infrastructure - suggestions
  • “Need to run APP, written in specific ver. Of PHP without rolling VM’s” (GAE flex)
  
  

Execute bash script on different hosts (ip list in file) via ssh

for host in $(cat /home/host_list | grep IP | egrep -v '^#' | awk '{print $3}' | sort);do

HostnameIP=$(echo "$host" | tr -d "[a-z]");
echo "##### $HostnameIP######";
sleep 4
ssh -i sshkey ec2-user@$host < bash_script

done

mongodb backup - bash

#!/bin/bash

DATESTAMP=$(date +%Y%m%d_%H%M)
LOG_FILE=/var/log/mongodb_backup/mongodb_backup.log
[[ ! -d /var/log/mongodb_backup ]] && mkdir /var/log/mongodb_backup
set -x
exec >${LOG_FILE} 2>&1

PID=$$
PID_FILE=/var/run/db_backup.pid
DB_HOST= DB_HOST_IP
DB_NAME='DB_NAME'
DUMP_PATH=/mongobck/tmp
DUMP_NAME=${DB_NAME}_${DATESTAMP}
S3_PATH='s3://mongodb-backups/'
LOGGER_TAG='mongodb_backup'

echo ""
echo "$(date): Starting MongoDB Backup"

[[ -f ${PID_FILE} ]] && (logger -sit ${LOGGER_TAG} "Instance of script already running. Exiting." && exit 99)
echo ${PID} > ${PID_FILE}

# Perform dump of mongo db
logger -sit "${LOGGER_TAG}" "Starting MongoDB backup of ${DB_NAME}"

/usr/bin/mongodump -h ${DB_HOST} --out ${DUMP_PATH}

logger -sit "${LOGGER_TAG}" "MongoDB Dump return $?."

# Compress mongo db
cd /mongobck/
tar -czvf ${DUMP_NAME}.tgz ${DUMP_PATH} && rm -rf ${DUMP_PATH}/*

# Copy compressed mongo db dump to s3
aws s3 cp /mongobck/${DUMP_NAME}.tgz ${S3_PATH} && rm -rf /mongobck/${DUMP_NAME}.tgz
logger -sit "${LOGGER_TAG}" "Backup of MongoDB completed successfully."

echo "$(date): MongoDB Backup Complete"
rm -f $PID_FILE
exit 0