vSphere Performance Tips And Tricks

Setting up your vSphere environment became easier over the last couple of year but a lot has changed from vSphere 4.0 to vSphere 6.0. In this article I will provide vSphere performance tips and tricks, which will be applicable to vSphere 5.5 and vSphere 6.0.

speedometer-148960_1280

 

  1. To get the best performance out of your vSphere Web-Client, use Google Chrome.
  2. Have vCenter hosted on a SSD or some low latency storage to have a pleasant experience.
  3. Ensure that Hyper-Threading is enabled, vSphere fully understands and uses it.
  4. Disable any power-saving features in the BIOS of your servers. Power-saving modes might cut the available compute resources.
  5. Use VMXNET3 driver on all your VMs. VMXNET3 is the only network driver which gets actively maintained by VMware.
  6. If you’re running 1GB vmnics, you might saturate the link and create a bottleneck. Monitor the average throughput on your vmnics and consider upgrading to 10GB.
  7. Be aware that there is no performance advantage of using RDMs, VMDKs or iSCSI direct-attached drives. Avoid RDM if possible. It is an administrative nightmare.
  8. When using iSCSI, make sure to configure port-binding properly.
  9. If you run jumbo frames, test the configuration. You would be surprised how many people mis-configure jumbo frames from end to end.
  10. For performance troubleshooting, look at ESXTOP. Here is a good article about it.
  11. Follow VMwares performance blog for up to date tips and tricks.
  12. Enable DRS & SDRS to utilize your cluster resources efficiently.

 

Note: If you feel I’m missing some important tips & tricks, please comment on this port or engage me on twitter and I’ll add your suggestions.

Ultimate Bash Guide For Beginners

bashEveryone who is new to scripting always doesn’t know where to start. There’s countless websites and blog posts out there with endless number of commands and bash guides. I don’t want to share all those super complex commands which are difficult to understand when you first start out with writing some scripts. Instead, I am providing a short bash guide for beginners. This post will cover the basics, some good manners and some more advanced commands.

One of the reasons why I started to write scripts in bash is because bash was just the language I knew due to Linux distros. However, I’m by no means an expert in writing scripts. If someone would ask me about my level of expertise in this area, I would rate myself with a “C”, maybe a C+.

Even though I am not an expert, I think it will be useful to share my experience and the commands which I have used most often. Below some general tips and commands which have helped me in the past and I still sue every day for some basic scripting.

General tips and tricks

Use clear and precise comments.
This will allow other users to easily read and understand your code.

#Validates that the variable is not an empty string. If it is an empty string it will set it false.
if [ -n "my-var" ]; then
    echo "Variable is not empty"
else 
    echo "Variable is empty"
fi

User proper indention
Indention will make your code easier readable. If your code is easily readable, it will be even easier to maintain it in the future.

if [ -n "$my-var" ]; then
    echo "Variable is not empty"
else 
    echo "Variable is empty"
fi

Let your script bail out if any command fails
For debugging of your script, you should use set -e at the beginning of your script.
Using set -e ensures that your script exists as soon as any of the lines is failing.

#!/bin/bash
set -e

Every bash script has to start with it
The first line in every bash script has to be #!/bin/bash
This will tell your terminal which interpreter to run.

Common Commands

Declare a timestamp function
For all my scripts I declare a function which stores the current timestamp when it is called.
I mainly use the timestamp variable for logging of events.

#!/bin/bash

timestamp() {
 date +"%Y-%m-%d_%T"
}

echo $(timestamp): "My script starts here"

a=5
b=6

if [ "$a" -gt "$b" ]; then
    echo $(timestamp): $a "is bigger than" $b
else
    echo $(timestamp): $b "is smaller than" $a
fi

The output will look like this:

./bs.sh
2015-04-15_22:06:55: My script starts here
2015-04-15_22:06:55: 6 is smaller than 5

Use $? to check the previous command
You can use $? to verify whether the previous commands returned 0 or 1.
0 = successful | 1 = failed

#!/bin/bash

timestamp() {
 date +"%Y-%m-%d_%T"
}

touch /tmp/test.log

if [ $? -eq 0 ]; then
    echo $(timestamp) "/tmp/test.log has successfully been created"
else 
    echo $(timestamp) "/tmp/test.log has not been created"
fi

If the script was able to create the file it will return:

./bs.sh
2015-04-15_22:14:04 /tmp/test.log has successfully been created

Read a file with comma separated values and act on it
When querying a database and saving the results as a CSV file, each results will be stored in a line and each value will be comma separated. Below is s simple script which scans a CSV file for the 4th value in every line and prints it, if it is not empty.

#!/bin/bash
timestamp() {
 date +"%Y-%m-%d_%T"
}

#This script expects a CSV file, called test.csv, in /tmp.

#This will read line by line till the end of the file

while read line; do 
#Echo $line, one line after the other, use "cut -d, -f 4" to look for values which are separated by "," and store the value in field "4" as $training

training="$(echo $line | cut -d, -f 4)"

#If $training is not empty, print $training

if [ -n "$training" ]; then
    echo $(timestamp): $training
else
#If $training is empty, print the line below 
    echo $(timestamp): "Field 4 was empty in file /tmp/test.csv"; fi; done < /tmp/test.csv