How to take care of sites during the holidays
The notorious human factor especially during long holidays is especially aggravated. In such weeks, by a multitude of excuses, establishments can open later, and business meetings can be postponed by mutual agreement. But what to do if you make a mistake for the welfare of your business? Or some important parts of your mortal body? Cloud monitoring services help to find peace of mind for those whose sites should work forever. The use of one of them in such situations will be discussed further.
Just yesterday, a complaint was received from an online store for unauthorized monitoring. Found a client, asked what relation to the checked site he has. It turned out the following: he made and paid for the order in this store, but then the site stopped working. The client could not even look at the support contact during the day! He found a simple solution - put the site on monitoring to find out when it will work again. And then forgot to take off. Question: why did not the store itself take similar steps?
So, there is a task: the site / server / service should work continuously for several days without human intervention. What can prevent this?
')
First, the usual failures that happen from time to time everywhere. Only here the night crash from Tuesday to Wednesday is solved by overloading / reloading from a backup on Wednesday morning. And on weekends there are frequent failures "from Friday to Monday." How much a site can be in such a case during holidays depends on the length of holidays of responsible employees.
Secondly, people also work at the hoster (for now ). And this means that they, too, can go on vacation, leaving on the farm 3.5 randomly broken hands. And besides, before going on vacation, they can easily plan an upgrade of the equipment / megaback / a bunch of updates, following the logic of "no one will notice." What to do with it - I wrote earlier .
Third, unscrupulous competitors or intruders often plan their black deeds for this particular time, counting on an increased reaction time on your part.
For example, here is how it happens when regular cleaning of a base falls off - how it grows (below) and how time requests to it (above) grow, and this is a light version of what can be in such cases.
All this leads to direct and indirect losses of money and customers and easily ruin the aftertaste after holidays.
The solution can be found for little money or even for free, depending on the needs. I will describe the approach using the HostTracker site monitoring service . I am ahead of the question: yes, there is Yandex.Metrica and similar things. But what are their differences and disadvantages, I explained in detail here . In short, this is not their main function, and sending an alert in a few hours is the norm. In addition, not all errors fall into their field of vision.
Here is one of the real examples of the past year. As you can see, the site is regularly not good, but on weekdays the problem is solved quickly. How long would the decision on May 1 take if not for the monitoring? Instead of 5 hours it could be a couple of days, and this is not uncommon.
In addition to the usual problems, sites during the long absence of their vigilant guards also like to catch other illnesses. For example, a domain or certificate may expire. Or he can please the DNSBL or Roskomnadzor lists. Or, as in the example above, the database will fail. And we will cure you of all this too.
As always, we welcome all suggestions and wishes. And have a good rest!
Real life
Just yesterday, a complaint was received from an online store for unauthorized monitoring. Found a client, asked what relation to the checked site he has. It turned out the following: he made and paid for the order in this store, but then the site stopped working. The client could not even look at the support contact during the day! He found a simple solution - put the site on monitoring to find out when it will work again. And then forgot to take off. Question: why did not the store itself take similar steps?
Holidays: waiting or afraid
So, there is a task: the site / server / service should work continuously for several days without human intervention. What can prevent this?
')
First, the usual failures that happen from time to time everywhere. Only here the night crash from Tuesday to Wednesday is solved by overloading / reloading from a backup on Wednesday morning. And on weekends there are frequent failures "from Friday to Monday." How much a site can be in such a case during holidays depends on the length of holidays of responsible employees.
Secondly, people also work at the hoster (for now ). And this means that they, too, can go on vacation, leaving on the farm 3.5 randomly broken hands. And besides, before going on vacation, they can easily plan an upgrade of the equipment / megaback / a bunch of updates, following the logic of "no one will notice." What to do with it - I wrote earlier .
Third, unscrupulous competitors or intruders often plan their black deeds for this particular time, counting on an increased reaction time on your part.
For example, here is how it happens when regular cleaning of a base falls off - how it grows (below) and how time requests to it (above) grow, and this is a light version of what can be in such cases.
All this leads to direct and indirect losses of money and customers and easily ruin the aftertaste after holidays.
What to do?
The solution can be found for little money or even for free, depending on the needs. I will describe the approach using the HostTracker site monitoring service . I am ahead of the question: yes, there is Yandex.Metrica and similar things. But what are their differences and disadvantages, I explained in detail here . In short, this is not their main function, and sending an alert in a few hours is the norm. In addition, not all errors fall into their field of vision.
Here is one of the real examples of the past year. As you can see, the site is regularly not good, but on weekdays the problem is solved quickly. How long would the decision on May 1 take if not for the monitoring? Instead of 5 hours it could be a couple of days, and this is not uncommon.
In addition to the usual problems, sites during the long absence of their vigilant guards also like to catch other illnesses. For example, a domain or certificate may expire. Or he can please the DNSBL or Roskomnadzor lists. Or, as in the example above, the database will fail. And we will cure you of all this too.
As always, we welcome all suggestions and wishes. And have a good rest!
Source: https://habr.com/ru/post/327548/
All Articles