Resource Abuse Policy
This policy is part of tems & conditions. Any website on our shared servers that uses a high amount of server resources (CPU, memory usage, and network resources) will be given an option to pay for a dedicated server. The abuse and misuse of shared system resources is a serious problem. Accounts on a shared server must share resources with other users so we must limit how much of the CPU and memory that can be used by any one account.
In all fairness, we cannot allow one or two clients to use all of the system resources on a shared machine and have all other clients suffer because of it. Excessive CPU & memory usage by one or more clients causes extreme slowness in all areas: MySQL, Email, HTTP and more. If the CPU & memory usage by a few clients gets out of hand, all sites hosted on the machine will return errors and not be accessible in any way.
Average Server Resource limit
Memory usage may not exceed 10% per domain/file/application
CPU usage may not exceed 20% per domain/file/application
Apache connections may not exceed 30 connections
15 MySQL maximum user connections allowed
350 emails per hour, per domain
There are special exceptions to these rules to be determined at the time we discover abuse. There are different combinations of usage that can create problems and other special cases. At any given time a server is unstable and an account is found to be the cause, we reserve the right to suspend the account to keep the server online and operating normally. If the usage is an increasing trend, we recommend the user invest in a dedicated server so the website can grow at a normal rate without restriction.
Suspensions
In the event your website has been suspended for resource abuse (abuse that cannot be resolved by removing/changing a script, etc.) and you do not wish to upgrade to a dedicated server to accommodate your site needs, we will only release the suspension within 24-48 hours. This is to ensure the CPU intensive site slows down in traffic to avoid further damage and problems. A backup will then be provided of your files or you will be allowed FTP/cPanel access for a certain time period. Any questions about our abuse policy may be answered by opening a ticket with our abuse division. You may visit the helpdesk by clicking here.
Inodes:
Accounts on any shared hosting account may not have an excessive number of inodes. Any account found to have over 100,000 inodes may be removed from our backup system to prevent overload. Any account seen to be exceeding 250,000 inodes is at risk for suspension. A warning will be sent and if no action is taken to resolve the issue, the account may be suspended. Every file on the account is counted as 1 inode. This includes mail as well as webpage content.
VPS Services
In order to control the VPS hardware node and to ensure that no single VPS account monopolizes the shared hardware resources of the VPS node (CPU/RAM) we use virtual limits to implement certain controls. The limits, or barriers if you will, which are most strictly enforced are the process limits (numproc), the RAM limits, and CPU usage limits. The 'numproc' limit which does not actually correlate directly with the number of processes you spawn, rather the number of kernel level threads any VPS account spawns is how the limit of kernel level processes are controlled. Please see below for the specified limits.
Numproc Limits
VPS1 - 140:140
We strictly limit the amount of RAM that each account is given. With any account, your account may only use the amount of RAM which is guaranteed to your VPS.
VPS Memory Limits
VPS 1 - 768MB
VPS 2 - 768MB
Our VPS offer an equal share of CPU to all accounts sharing the hardware node. With this given, the CPU limits are as follow:
CPU Usage Limits
VPS1/2 - 700MHz Guaranteed, burstable to 850MHz
Any single VPS account which is surpassing the above limitations is considered "Abusive". Should your VPS be abusive for more than 12 hours you will be at risk of suspension. Nepal Domain Hosting reserves the right to suspend, without prior notice, any VPS account which is causing the main hardware node to crash.