6 Common

Causes of
Poor Database
Performance
. . . and what to do about them
 6 Poor Database
Common
Causes of

Performance
. . . and what to do about them

Your company relies on databases to succeed in business.

Whether you provide gaming services, software as a service, or an
e-commerce site, you need a fully optimized, highly available, and
superbly performant database to ensure your applications meet
customer expectations.

When even the slightest slowdown can frustrate users, damage

your brand, and lead to revenue loss, you must be able to quickly
identify the true causes of database performance problems and
fix them fast.

Below are the top six causes of poor database performance

that you should look out for.

2 | Six Common Causes of Poor Database Performance

 1
: Inefficient Database Design
How you design your database environment impacts everything you do down the line.
Poor design decisions cascade throughout the life of your applications, potentially leading
to serious problems with performance and scalability. The cost to you and your team will be
high as the volume of complaints grows and you find yourself applying valuable resources
to retroactively fix early flaws.

Why is good design so important? Because in complex database infrastructures, everything is

connected. Applications, which once comprised a single stack, now encompass many diverse
services. Databases work within large ecosystems of components. And infrastructures stretch
across localities, countries, and continents, operating across different environment types,
including cloud, on-premises, and hybrid.

With all these interdependencies, it’s important to ask good questions as you consider the
design of your database infrastructure. Here are a few:

Capacity. Understanding how Cloud. Although cloud vendors

data flows through your architec- manage operational automation
ture helps you create the most and day-to-day tasks, you still own
effective environment for your applications. ensuring the performance and availability
How much data will you store and where? of your database. How will you manage
How quickly will it need to be accessed? and monitor queries? How will you handle
How many concurrent users do you disaster and recovery? How will you ensure
anticipate for various applications? your database is secure?

Workload. To optimize your Scalability. As data grows,

infrastructure, you need to your database slows down. How
forecast workload. Will it be high- will you scale as data grows? Can
read, high-write, or both? Will it be steady, you partition or “shard” your database?
or will it change based on specific events? Can you create database clusters?
Does your database engine match your
workload type? Limitations. Infrastructure
teams support multiple applica-
Data types. The more data tion stacks. Do you have the power
you store, the less you can fit into to pick optimal configurations, or is that
memory, so your data types dictated by your company or other teams?
must be as efficient as possible. Will they
increase the footprint of the data they’re Support. Think about what will
storing? Will they make processing and be required of your team. Do you
handling data more efficient? have the internal knowledge to
manage your database? Will you need ex-
Software ecosystem. Data ternal help to ensure all your components
bases don’t exist in isolation. What work together now and in the future?
other software will you need along
with your database? How will you integrate
with third-party plugins? Will data be fed
from third-party applications?

3 | Six Common Causes of Poor Database Performance

2
: Errors in Deployment
Even a well-designed, well-implemented system must be deployed correctly. A lot of
deployment issues are trivial: using the wrong hardware or the wrong instance type in
the cloud, using spinning disks instead of flash-based storage, or using an inappropriate
amount of memory. Typically, these types of issues are caught quickly and are easy to
resolve. But some are not; for example, database change errors, which can cause serious
problems in production and are time-consuming to diagnose.

Thorough testing helps prevent deployment issues, but problems can still be missed.
Before you go to production, deploy your database in a test environment first and make
sure you have a backup of your data. Also, implement version control of your database and
schemas so you can roll back if something goes wrong and easily trace who did what and
when. Traceability allows you to more quickly isolate and fix issues.

More than the database, network-related issues are usually the biggest cause of deploy-
ment related problems. Even in the same data center, adding multiple switches or slow
firewalls between the application server and the database can multiply latency.

Finally, if you place the database server and application/web server in different availability
zones or data centers, the latency increase gets even more dramatic. As you decide where
your data will sit, keep in mind that data needs to be near running applications to minimize
latency and provide a highly performant user experience.

3
: Software Misconfigurations
The configuration of software components is important to database performance.
Just one misconfiguration can not only slow your applications, but cause downtime,
data corruption, data loss, and security breaches.

From the application standpoint, a misconfigured driver is a common database-related

issue. Configuring a Java connection pool to be too small, for example, can cause low
performance and application errors. Setting it too large can lead to connection errors,
poor performance, or even crashes. Setting the wrong isolation mode in the connection
settings without understanding the implications, can cause the application to work
incorrectly and potentially cause data corruption.

4 | Six Common Causes of Poor Database Performance

4
: Over Indexing
The difference between your database being fast, responsive, and scaling properly often
depends on how effectively you use indexes. The big advantage of indexes is that they
permit queries, updates, and deletes to run as fast as possible. However, over indexing
can lead to massive slowdowns in overall performance.

To be really effective at improving performance, indexes should be as small as possible.

In addition, the number of indexes you need depends on your application workload. Some
developers create a lot of indexes without a specific reason or to test a query, but then
forget to drop them, which, in some cases, can make the size of the indexes larger than the
data itself. The general rule is to create all the indexes your application really needs
for solving the most frequent queries—not one more or one less.

Creating indexes for solving queries is a good habit, but you need to use them wisely.
Consider maintaining your indexes from time to time, dropping all duplicates and unused
indexes that may slow down performance.

5
: Resource Saturation
Database systems have only so much memory, CPU, disk IO, and network resources,
and can’t do more than their constraints allow. Optimizing your software configuration,
schema, and queries places less demand on these types of resources. For example, a web
application developer may write code that involves several separate queries instead of
batching them, which unnecessarily increases load on the CPU and memory.

Designing for efficient resource usage is important, but so is having a way to continually
monitor that usage so you can address problems as they arise. Consider using a monitor-
ing tool that provides features that allow you to analyze your database queries and identify
and fix issues as quickly as possible. It should also provide well-designed dashboards
that give you easy insight into system metrics, such as CPU, network, memory, and
disk performance.

5 | Six Common Causes of Poor Database Performance

 6
: Database Misconfigurations
When users start complaining about application performance, how do you know whether
it’s the database or something else causing performance degradation? Sometimes finding
the answer is simple. For example, you check the database process list to see which appli-
cation queries have been running for more than 30 seconds. But in other cases, it might not
be so obvious, like a misconfigured buffer pool or log file size.

Here again having a tool that gives you deep insight into what’s happening in your data-
base system allows you to determine whether the problem is with the database, and if so,
identify what it is. This should be a database-focused tool for in-depth troubleshooting and
investigation so you can fully understand the health of your infrastructure.

It should include capabilities for monitoring the state of your systems based on predefined
sets of metrics. It should also provide observability, which gives you deep visibility into the
internal states of your systems, helping you actively discover the “unknown, unknowns.”

Finally, a database brought down by a cybersecurity attack due to misconfigurations does

not perform at all. Make sure you have a tool that helps you quickly find and fix common
data security issues across all of your open source databases. It should provide daily secu-
rity checks for common issues, alerts for when non-compliance is found, and the ability to
audit your security check history.

How Percona can help

Percona can help you determine whether your database issues are due to one of the
reasons above or something else entirely. We’ll assess your current deployment and
provide a detailed analysis and plan for optimal performance, which you can implement
yourself or with our help.

LEARN MORE about Percona Database Consulting.

Want to get started right away? Contact us at [email protected].

About Percona
Percona is widely recognized as a world-class open source database software, support, and
services company. The organization is dedicated to helping businesses make databases and
applications run better through a unique combination of expertise and open source software.
Percona works with numerous global brands across many industries creating a unified experience
for monitoring, managing, securing, and optimizing database environments on any infrastructure.
Percona equips organizations with the freedom to choose, the freedom to create, and the
freedom to make a difference - helping them scale and innovate with speed as they grow.
For more information, visit percona.com.

Databases run better with Percona.

For more information please contact us at:
+1-888-316-9775 (USA), +44 203 608 6727 (Europe),
or via email at [email protected] © Copyright 2023 Percona® LLC. All rights reserved.
170123