Of course, we monitor our servers and the hosting packages they hold for you. To ensure websites can work smoothly, we keep an eye on the main points, for example disk space, traffic and RAM. If there is impact on the health of your website, we will e-mail you accordingly.
Below, the categories on which our proactive monitoring keeps an eye and reports are explained. In this manual, we explain them further and give tips to get the root cause and solve any issues related to them.
- Disk space
- Traffic
- Processing power (CPU)
- Work memory (RAM)
- Memory usage
- Memory issues
- Connections (Entry processes)
- Used connections
- Processing (nProc)
- Used processes
- Process issues
- I/O - traffic
Disk space
What is it?
A self explanatory category is the disk space a hosting package uses. At the moment when your maximum limit is reached, you cannot store any more data on your hosting package.
Symptoms and problems
When a hosting package reaches its disk space limit, then no new data can be stored on the package. This means that no new e-mail can be received, or that the website has issues when storing data, for example with session files or uploads.
Possible causes
High disk space usage can have multiple causes. It can be that for example, e-mail is stored for a long time within the e-mail accounts, after which they get fuller and fuller. Sometimes, there is other (unnecessary) backup data stored on a hosting package, which takes up a lot of space. Other causes which are more due to misconfiguration can be overflowing cache folders within a web application, or extremely large databases.
Solution
The solution is closely related to the cause of the high disk space usage. When a lot of e-mail is being stored, consider removing older messages. If there are unnecessary backups on the package, consider downloading them and storing locally to free up space on the hosting package.
If do not want or are unable to remove data, the alternative is to upgrade your hosting package to an alternative with more disk space.
Traffic
What is it?
The traffic is the amount of traffic going to and coming from a webhosting package on our servers. You can view this in the same manner as the data limit on a mobile phone plan.
Symptoms and problems
When a hosting package has reached its monthly allotted traffic, then the package is temporarily blocked until the first day of a new month. This prevents the traffic from going above the set maximum, but also means that the website cannot be reached anymore by visitors.
Possible causes
High traffic can have multiple causes. It can be simply because the website has been visited a lot, and that high visitor numbers therefore cause more traffic. Other causes can include indexing done through search engines, external references to a file or a part within the website. In some cases, misconfiguration, such as a very enthusiastic e-mail program or part of the website, can also cause way more traffic than usual.
You can find the cause more directly within DirectAdmin by going to 'System Info and Files', then 'Site summary / Statistics / Logs'. Under 'User log' in the top right, and then clicking the month which you would like to see, is the traffic specified per part of the hosting package (for example, the website, e-mail or FTP).
Solution
When the traffic for a certain month has been reached, you need to wait until the new month arrives. That is of course not desirable, because the package is then unreachable for the remainder of the current month. You can therefore upgrade your hosting package (temporarily is also a possibility) so the website can be reached again.
CPU usage
What is it?
The CPU is the working horse of the server. Everything a server does, is calculated within this central processor. When using our server, you as an user get your own share of this CPU. Packages can contain multiple CPU-cores, allowing them to have their own processing power.
Symptoms and problems
When scripts on the website, for example an application or CMS, take up a lot of processing time of the CPU, it can be that the CPU is fully maxed out. That in and of itself does not always have to be a problem, but when new processes arrive and need to wait for CPU, it can cause delays and processes can stack as a consequence.
Possible causes
CPU problems are relatively rare. It is usually the case with extremely inefficient or heavy scripts, or the usage of for example an outdated PHP version which can handle resources less efficiently.
It is also possible that the website is correctly functioning and fully optimised, but is simply visited a lot that there is a need for more CPU.
Solution
The first thing to check is whether the PHP environment of the hosting uses the latest available PHP version. In later versions of PHP, server resources are handled significantly better. Make therefore sure your hosting package is using the latest version.
If you are already using that, a good next step is to check out the processes which are active when there is a peak tax on the CPU. You can do so through SSH by using the command ps aux to get a proces list.
You can also do this by checking the server logs, to see which requests are being done to the website. In some cases, this can be an indication of which parts within the website uses a lot of CPU. You can find these within DirectAdmin by going to 'System Info and Files', then 'Site summary / Statistics / Logs'.
Next to that, there are other steps you can take to lower the resource usage. Think of optimising your website by updating your CMS to the latest version, set up caching, or delete unused or heavy plugins, modules, extensions, and themes.
Is the hosting package fully optimised but CPU usage remains a problem, consider upgrading your hosting package to one which offers more CPU cores.
Work memory (RAM)
What is it?
When referring to work memory, the physical memory is addressed on which the processes on the server run. You have your own guaranteed amount of RAM available on your hosting package, which is used for the processes of your hosting package.
Symptoms and problems
When a script on the website, for example in the shape of an application or CMS, is using extraordinary amounts of work memory, it can be that there is no memory left on your hosting package. This poses a problem directly, since there is no memory for processes to use any longer. In almost all cases this results in an error message within the website, for example a 500 or 503 error.
Because full usage of the work memory is of direct consequence of the availability of your website, we monitor this as well.
Possible causes
Excessive use of RAM has multiple causes. There can be a misconfiguration such as an extremely inefficient and heavy script, a heavy plugin, or an unoptimised CMS is demanding too much memory from the server.
It can also be that an application functions correctly, but simply has higher memory demands. Magento is for example known for asking more RAM from the server.
Additionally, and the same goes for CPU, an outdated PHP version can influence in a negative manner the way resources are handled.
Lastly, a website can also be simply successful or popular enough that it requires more RAM.
Solution
The first thing to check is whether the PHP environment of the hosting uses the latest available PHP version. In later versions of PHP, server resources are handled significantly better. Make therefore sure your hosting package is using the latest version.
If you are already using that, a good next step is to check out the processes which are active when there is a peak tax on the memory. You can do so through SSH by using the command ps aux to get a proces list.
You can also do this by checking the server logs, to see which requests are being done to the website. In some cases, this can be an indication of which parts within the website uses a lot of memory. You can find these within DirectAdmin by going to 'System Info and Files', then 'Site summary / Statistics / Logs'.
Next to that, there are other steps you can take to lower the resource usage. Think of optimising your website by updating your CMS to the latest version, set up caching, or delete unused or heavy plugins, modules, extensions, and themes.
Is the hosting package fully optimised but memory usage remains a problem, consider upgrading your hosting package to one which offers more RAM memory.
Connections (Entry Processes)
What is it?
The webserver is responsible for serving the website to your visitors. Webservers have a maximum amount of connections they can handle at the same time. To guarantee that websites do not hinder each other, every user has the same number of entry processes, which is 40.
Symptoms and problems
The problem is born when the maximum number of processes for a website has been reached. There are no new connection 'slots' available for the user, which will result in a 508 error for a new request. The website is then, for the visitor who is the 41st connection, unreachable.
Possible causes
Remarkably, when the maximum number of processes have been reached and the problems resulting thereof, it is more often than not, not a sole problem. Reaching the number of entry processes is the result of another problem. If a package for example hits the CPU or the I/O traffic limit, it can result in a stack of the number of connections at the webserver, since the processes are waiting for available resources to be able to run. Then, the hosting package uses all connections, which results in errors.
Solution
It is a good first step to check which other resources have also reached their limit when the entry processes limit has been reached. If the package is using the CPU simultaneously, or if there is maximum I/O traffic, it is very likely that this is the cause of the entry processes limit.
When that is not the case, a different cause should be find. There can be a problem with the code within the website which results in requests taking longer than necessary, which causes them to stack.
You can then best check the server logs. You can do so through SSH by using the command ps aux to get a proces list.
You can also do this by checking the server logs, to see which requests are being done to the website. In some cases, this can be an indication of which parts within the website uses a lot of memory. You can find these within DirectAdmin by going to 'System Info and Files', then 'Site summary / Statistics / Logs'.
Finally, there can also be abuse; for example a spammer, or a brute force hacked, which performs repeated requests to a certain part of the website. You can then consider blocking them based on their IP-address through your .htaccess file.
Processes (nProc)
What is it?
The total number of processes from your hosting package is also being monitored. A webhosting package can run 100 simultaneous processes maximum. The available 100 process 'slots' differ from the 40 entry processes with the webserver, in such a way that a proces can also run on the server itself without a reference to the website. These are processes such as cronjobs, a script done through SSH or which started on the background through the website.
Symptoms and problems
The problem is born when the maximum number of processes on the webhosting package has been reached. There are no new process slots for that specific user, which is why starting a new process results in a 503 error. When this process has been started from the website, the website itself is then unreachable for the visitor who tries to start the 101st process.
Possible causes
It is pretty rare that the number of processes stack to 100, since in most cases, these processes are started from within the website, and will run therefore earlier into the maximum number of entry processes with the webserver. It can be however, when for example a proces has been started from the website that keeps running in the background. It can also be that a (wrongly configured cronjob) starts a process, which keeps on running significantly longer than intended. The cronjob can then take in on itself, resulting in the stacking of a number of processes as a consequence.
Solution
You can best check the server logs. You can do so through SSH by using the command ps aux to get a process list.
You can also do this by checking the server logs, to see which requests are being done to the website. In some cases, this can be an indication of which parts within the website uses a lot of memory. You can find these within DirectAdmin by going to 'System Info and Files', then 'Site summary / Statistics / Logs'.
Once you have identified the cause of the stacking processes, you can set the right steps in order to solving them, by for example changing the scripts responsible or the cronjob.
I/O - traffic
What is it?
I/O traffic is also known as the bandwidth your hosting package has in use for reading and writing operations on the file manager. This is measured in MB/s (megabytes per second).
Symptoms and problems
When a webhosting package is using all the available I/O traffic, this means that processes for this package cannot read or write any faster to and from the file manager. This can also mean that processes which need to wait until bandwidth is available to finish running. The consequence is delay within the waiting processes.
When this is happening often, then it will lead to stacking the processes (for example, with the CPU), and the package can get in trouble with for example the number of simultaneous connections or processes.
Possible causes
I/O traffic is determined by edits on the filing system. I/O problems occur when a package, for example due to misconfiguration, generates an extreme amount of reading and writing traffic. This can be due to caching which is not configured correctly and keeps on writing new data, or the usage of extremely large images on a page which is frequently visited. Additionally, inefficient backup or migration plugins can cause I/O traffic issues.
At play here as well is the PHP version which is being used. Using an older PHP version can lead to (unnecessary) high I/O traffic.
Solution
Most websites do not need more than 1 MB/s for I/O traffic. Our packages do offer more than that. If your hosting package is therefore constantly reaching the limit for the I/O traffic, it is good practice to see where it is coming from. Do check for example any caching functions within your website, and check which reading and writing operations are being done upon each website visit. It is also beneficial to check whether you are using the latest PHP version.
Browser-caching, so saving parts of the website in the browser of a visitor, is very efficient in reducing I/O traffic. When parts of the website are stored locally with the visitor, these do not have to be loaded from your website and therefore also not from the file manager.
Files (inodes)
What is it?
One inode is equal to one file on your hosting package. Because disk space can impact your speed negatively, we have set a maximum to the number of inodes. After logging into DirectAdmin, you can see how many inodes you are using. You can also use the following SSH command for that.
find . -type f | cut -d/ -f2 | sort | uniq -cSymptoms and problems
When the inode limit has been reached, you can no longer write off files in your hosting package. It is possible that you reach the limit when a misconfigured caching plugin is writing off and placing a bunch of files on your package. Once the limit has been reached, you can for example no longer upload no new images.
Possible causes
High inode usage can have multiple causes. It can be a misconfiguration such as caching within a web application, or a large amount of media files still present on the package.
Solution
First, determine the cause. Where is the high amount of inodes coming from and is the usage of that many files legitimate? If there is a misconfiguration, you can solve the problem by removing the files which are no longer needed and adjust the configuration.
If the usage is legitimate, you can consider upgrading your hosting package. Investigate in that case in which package the limit is enough for you.
Checking the resources within DirectAdmin
You can check for yourself which specific resources your hosting package is using. You can check this under 'Resource usage'.
- Log into DirectAdmin.
- Within the menu, go to 'Extra functions' and then to 'Resource usage'.
3. Click on 'Details' in the next screen. Then, you will see an overview with different graphs. You can see here when, and what specific parts of the resources are causing issues. At the bottom there is a complete table with values which were measured on specific time stamps. You can also change the period here, up until 30 days in the past. Standard, this is set to the past 24 hours.