In this article, we explain how CloudLinux limits work on your reseller hosting with Etheron, and how you, as a reseller, can set the size of your packages yourself using CPU, RAM, I/O, IOPS, EP and NPROC. This way you know exactly what each setting means and how to tune it properly for your customers.

What is CloudLinux?

All of our webhosting and reseller servers run on CloudLinux. CloudLinux ensures that every hosting account runs in its own isolated environment (called an LVE, Lightweight Virtual Environment). This means one busy or heavy website can never bring down the entire server, or your other customers.

As a reseller, you decide for yourself how many server resources each of your customers (or packages) may use. You configure this through the DirectAdmin control panel, under CloudLinux Manager.

Where do I find this in the panel?

1. Log in to your reseller account in DirectAdmin.
2. Go to Reseller Level (or Admin Level if you have full access).
3. Click Extra Features.
4. Open CloudLinux Manager.

Here you will find three important tabs:

• Packages: here you set the limits per hosting package. Every user on that package automatically gets the same limits.
• Users: here you see the current usage per user, and can set an individual exception if needed.
• Current Usage / Statistics: here you see live how much of their limits a user is currently using.

Note: Preferably work through Packages instead of adjusting individual users. This keeps it clear which customer is on which package, and prevents you from drifting away from your own package structure.

What do the different limits mean?

CPU / SPEED

This is the processing power an account may use, expressed as a percentage of a single CPU core.

• 100% equals 1 full core.
• 200% equals 2 cores, and so on.

When this limit is reached, the website is not shut down, but throttled: the website runs slower until there is room again.

RAM / PMEM (memory)

This is the amount of memory the account may use, for example 1GB or 2GB.

If an account goes above this limit, CloudLinux refuses to allocate extra memory. This usually results in an error such as a 500 or 503 error, or an "out of memory" message, for example in WordPress.

I/O (disk speed)

This is the maximum speed at which files may be read and written, expressed in KB/s or MB/s.

When this limit is reached, the account is throttled, not stopped. A too low I/O limit is often noticed during heavy actions, such as a backup, an update, or processing a lot of images.

IOPS (disk operations per second)

This is the number of read and write operations an account may perform per second. This is especially important for websites with a lot of database traffic (for example webshops).

EP (Entry Processes)

This is the number of simultaneous connections an account may have to, for example, PHP, SSH or cron jobs. Think of it as the number of "simultaneous visitors requesting something dynamic at the same time".

If this limit is exceeded, new visitors receive an error such as 508 Resource Limit Is Reached, until room becomes available again.

NPROC (number of processes)

This is the total number of active processes an account may have at the same time, for example PHP processes, SSH sessions or cron jobs. In some panels and documentation this is also referred to as MPROC, which is the same thing as NPROC.

When this limit is reached, no new processes can be started until an existing one stops.

Note: Always set NPROC comfortably higher than EP, as a rule of thumb at least EP + 15. If NPROC is too low relative to EP, normal processes (such as Apache itself, SSH sessions, etc.) can already hit the limit.

How do you configure this correctly?

Guideline starting values per package size:

Limit	Small package	Medium package	Large package
CPU / SPEED	100%	200%	300%
RAM / PMEM	1 GB	2 GB	4 GB
I/O	5 MB/s	10 MB/s	20 MB/s
IOPS	1024	2048	4096
EP	20	30	50
NPROC	100	150	200

These values are a good starting point, not a fixed rule. Adjust them based on the type of websites your customers run.

• Only a simple or static website? The default values are usually more than sufficient.
• WordPress, WooCommerce or other heavy CMS sites? Set RAM and EP a bit higher, these are the limits dynamic sites hit first.
• Webshops or sites with a lot of database queries? Pay extra attention to IOPS and I/O.
• Backup or migration work? A low I/O limit often causes slow or stuck backups, even if the site itself is not heavy.

Note: Never set limits unnecessarily low. Limits that are too low cause slow websites, error messages, and unnecessary support requests, even when the website itself is not doing anything wrong.

Frequently asked question: my customer is getting errors, what now?

If a customer regularly hits their limit, check the following:

1. Which limit is exactly being hit. You can see this in CloudLinux Manager → Users or in the statistics.
2. Whether the problem lies with the website itself. Many isolated EP or CPU faults with no clear cause can point to a missing cache, a heavy plugin, or even bots hammering the site.
3. Whether the limit is simply too low for what the customer needs. If so, raise the package or the limit of that specific package, instead of continuing to set individual exceptions per user.

Troubleshooting

• Website loads slowly or gives 500/503 errors: check RAM (PMEM) and CPU first, these are the most common causes.
• Visitors get "508 Resource Limit Is Reached": the EP limit has been reached. Check whether this is normal traffic or, for example, bot traffic, and raise the EP limit if needed.
• Backups or large file operations get stuck or take extremely long: raise the I/O limit.
• Webshop or database-heavy site is slow: check the IOPS limit.

Can't figure it out, or would you like help setting up your reseller packages? Feel free to reach out via Discord or create a support ticket.