Search our site for answers
Help Docs Server Administration Linux Server Administration Getting Started with Cloud Linux

Getting Started with Cloud Linux

CloudLinux is a control manager for your Linux server that allows you to partition, allocate, and limit server resources like memory, CPU, and connections on a per account basis. But how do you know what limits are appropriate to set for each account?
Getting Started Server Management
3 min read

CloudLinux is a control manager for your Linux server that allows you to partition, allocate, and limit server resources like memory, CPU, and connections on a per account basis. To learn more about CloudLinux, see What is CloudLinux?

But how do you know what limits are appropriate to set for each account? The answer is generally complicated and it depends on a number of factors like how much traffic a user regularly receives, how busy their database is, and how quickly their site changes. However, there are some baseline settings that you can start with that should help to get you started. You can always tweak the settings in the future if individual client needs are different from the standard configuration.

Here are some of the main resources you would use CloudLinux to restrict, along with recommended values for each:

  • Memory

CloudLinux is able to identify, in real time, the amount of memory actually used by an end customer’s processes. Physical memory limits are especially effective in preventing out of memory (OOM) issues and customers’ ballooning memory usage, which destroys caches and causes server overload.
Depends on the number of accounts and the amount of RAM available on the server. Divide total available ram by total accounts to find a maximum.

  • IO

IO limits restrict the data throughput for the customer. They are measured in KB/s. Because IO is one of the scarcest resources in shared hosting, the ability to put an upper limit on customer use is vital.
5 MB/s is a good baseline

  • Number of Processes

Number of processes limits control the total number of processes within LVE. Once the limit is reached, no new process can be created until another one has finished. This effectively prevents fork bombs and similar DoS attacks.100 is a good minimum

  • Entry Processes

The best way to think about this type of limit is as the number of web scripts that can be executed in parallel by visitors to a site. These limits are important to preventing single sites from hogging all Apache slots, thus causing Apache to be unresponsive.Start with 10 (which allows 10 concurrent users per second). Increase as needed

  • Inode

An inode is a data structure on a file system that is used to keep information about a file or a folder. The number of inodes indicates the number of files and folders an account has. Inodes limits work on the level of disk quota.Unlimited – generally not an issue if other resources are well-balanced.
Was this article helpful?