Kamil Porembiński
Server failure is one of the worst scenarios that can happen to a thriving business. Websites, applications and other resources put on the same server stop working. How to deal with breakdowns? Is it always possible to identify the cause of failure? Is there a chance to eliminate them completely?
Each of these questions is answered in this article.
Let me explain it to you with a very simplified example.
Meet Janek. Owner of a successful online health food store. Some time ago he gave up hosting. He chose a dedicated server with much higher parameters so that it could handle a much larger number of clients. He understands something about IT, so he didn’t need anyone else besides the developer handling the site design. With the recent fair, he met a lot of influential people. As it turned out, some of them advertised the store at the same time, due to the excellent quality of the products. Janek sees the statistics in Google Analytics growing. He is pleased that more and more people are finalizing their orders. Then, all of a sudden, everything drops to zero. The page displays an error. Horrified, he calls the developer, who can’t diagnose the problem and throws in at the end of the call that it’s definitely something by the server provider*.
Janek calls the company where he bought the resources. There he learns that it will take several hours to diagnose the problem. Zalamy is trying to fix the problems himself. He is looking for instructions on the Internet. First search and we have this. This is probably due to too much load on the server. Janek follows the instructions and enters the first commands. The service is still unavailable. In addition, he sees that there is nothing in the files on the server. He wants to undo the changes, but can’t. Devastated, he calls the server provider again. The administrator checks the infrastructure and says the service is empty and sees no backups. Janek was left with an empty server. He lost all the files he had been working on for several years.
First, the increased traffic itself does not immediately cause the server to fail. It causes a heavy load that the server resources could not handle. The server, after most users left, would return to normal and you could say that nothing happened. However, Janek’s further actions led to massive problems.
Second, backup is a must. Without it, under no circumstances should you change anything on the server. One misspelled command can delete all files.
Another thing is to make changes without IT’s knowledge. It’s like letting an IT person into a beauty salon to do a Kobido treatment ???? Therefore, let’s not touch things that are totally foreign to us.
Fourth, the right person in the right position. The developer designing the site will not always be familiar with maintaining it. These are completely different areas. In addition, if you haven’t purchased the service from third-party providers, you can’t count on immediate assistance. Mainly due to the lack of knowledge of your environment.
It is also worth paying attention to the preparation of servers. If you expect at least a few percent more traffic, be prepared for it. Users don’t like to wait, so recommended in such cases are performance tests that will check the strength of the site.
In summary, the failure of servers can be interpreted very differently. However, we can put it as longer unavailability of services for most users.
The sites are not working for various reasons. The most common cause of failure is:
I am no longer going into the timing of the problem itself here. These are the general reasons causing inaccessibility. It is worth noting that there may be many more. Even more so if you divide them by the system and the hardware in the server room itself.
I will address one of the reasons. No response on the customer side. As you know how it works, you don’t have to be too interested. When it stops that’s when we are most engaged. However, a common cause is precisely the lack of response. One of the tasks of the Administrator is to anticipate errors and communicate to the client the available solution. Without confirmation of the changes, we can do nothing, which later indicates the cause of the failure.
Step one. Get help.
If you are not familiar with IT operations, contact the people responsible for the servers. Whether it be an outside company, your employee or the support of the company from which the server resources were purchased. If this is the first time and you do not know how to deal with this situation, then write to us. We will help you as quickly as you allow.
Step two. Determine the cause.
Try to determine the cause of the failure. Using all available repair methods you will only lose time, which in this case is very valuable. First of all, check the messages available in the panel with servers. It may happen that failures are caused by changes being made at the supplier or saturation of the site, which can be seen directly from the command center.
Step three. Solution to the problem.
All depending on your IT skills. If they are not sufficient, wait to contact the Administrator.
In case you feel up to fixing the problem, remember to make sure you have a backup before proceeding. If not make one. The next point is to proceed to solve the problem.
It is impossible to include in one article all the instructions for solving server problems. That’s why I’ve pointed out the basic principles of emergency response. I will add one more important rule. Don’t panic. Stay calm and try to find the source of the problem.
It sounds beautiful, but as we all know, any infrastructure is prone to failure. The guarantee of an efficient server is first and foremost adequate support, which will monitor your services around the clock. You have access to a wide variety of solutions. Remember to always analyze each of them and choose the ones that will give you the greatest accessibility and efficiency.
Qlos’ standard is communication to learn about our customers’ needs. We talk about them and choose the right server administration package. We do not offer unnecessary add-ons that only generate costs.
I invite you to join the conversation. Whether through a form or via Linkedin. You will gain security, time and “peace of mind” !!!!
