A store’s performance can take a (temporary) dive for many different reasons. The store may have reached the limits of its hosting space, the Magento processes have gone haywire, or a recently updated extension is causing issues. In practice, any store will encounter such unanticipated obstacles at some point. That’s why it’s good to know that even if you won’t always notice, Hypernode is there to support your store.
Our Hypernode platform has various smart features that prevent your store from freezing in case of performance issues. Most of the time you won’t even notice it as we handily save you from issues you didn’t know you had. So, we thought we should tell you a bit more about them.
#1 Automatic recovery
A Hypernode may go offline or stop responding. This can happen for various reasons. For example, there are more store visitors than the server resources can handle. Or processes in the Magento application have gone haywire and are causing issues. These aren’t rare occurrences. They affect stores on our platform multiple times a day. It’s unavoidable for a web shop with many different functionalities, tasks and processes.
The solution itself isn’t that hard. Often, all that’s needed to get the Hypernode back up again is to shut down or restart the processes. What’s more challenging is to achieve this as quickly as possible, with minimal downtime. Hypernode automates such recovery operations, putting us way ahead of other Magento hosters.
Whenever our script detects that not all Hypernode services (PHP, NGINX, MySQL) are active, our monitoring system is alerted instantly. Our smart monitoring soon ascertains which specific service is having issues and restarts this – completely automatically. If this doesn’t have the desired effect, it is immediately succeeded by an automatic restart of the specific Hypernode. In at least 90% of cases, this is enough to solve the issue. Only if restarting proves insufficient a human assistance will be enlisted. A technician will then receive an alert that manual investigation and resolution is required. Automated response times are always faster than humans. The entire process takes ten minutes at most. That means your online store can be back up and running three times as fast as with a non-automated recovery process. As you can imagine, the advantages of this method are even greater outside normal working hours, when the technician won’t always be behind their computer.
#2 Out-of-Memory event
The instant a Hypernode’s entire available memory is in use, automatic emergency procedures will activate. Non-essential processes come to a halt, ensuring that the critical processes (which are keeping your store online) can continue to run. After such emergencies you’ll receive an email, so that you’re aware of the intervention and can investigate the underlying cause behind the high memory usage.
#3 Log rotation
Log files contain lots of information. That can be very useful, but without regular deletion of old logs, the file will grow unwieldy and start to hog an unnecessarily large amount of disk space. At Hypernode, we employ daily log rotation to restrict the size of the log file. Outdated logs are deleted automatically. We do this for many different logs.
#4 Emergency log truncation
For NGINX and SFTP logs, we even do a little more. Sometimes an error in the store’s code can create so many logs in rapid succession that the log file reaches extraordinary proportions. Log rotation alone is no longer sufficient, as this only happens once a day. To prevent NGINX and SFTP log files from filling up the Hypernode disk completely and crashing the store, we automatically delete logs when the log file becomes too large. If your Hypernode requires this emergency intervention, we’ll also send you an email. That way you’ll know immediately that something at your store is in need of attention.
Automated response times are always faster than humans.
Unique stability and reliability thanks to standardized setup
Automating as many tasks and processes as possible saves valuable time and prevents store downtime due to unanticipated performance issues. This is only possible thanks to our standardized setup. Something we firmly believe in after twenty years of hosting experience. With standardization, all Hypernodes have the same setup. This setup still gives you or your technical manager plenty of freedom (you’d be surprised), while also ensuring that we can use the above mentioned processes to provide unmatched stability and reliability.