A server crash can be devastating. Luckily, this is not some inescapable inevitability – but how do you prevent it? And then, if one does happen, how do you recover?
What is a server?
A server is a computer that provides information to other computers on a network in the strictest definition of the word. This information can be a lot of things – in the gaming world, a server is a computer supplying a connection point between gamers, so they can all play on the same map together even though they may be all over the world geographically. Entertainment services like Netflix or Youtube are storing their massive video libraries on servers, so users with accounts can access them on demand instead of having to download them.
In the business world, servers can certainly still provide data to a continent’s worth of clients, but even smaller businesses may benefit from having one spot with all the files accessible on it. A common pain point, for instance, is scattered accounting documents. Having them all on a single server, with one receipts folder that’s updated and checked regularly by the person doing the expenses, makes this pain point a non-issue. Localizing a client’s data onto a server instead of giving multiple people redundant copies of the same things means updating their data is not only easier and more efficient, but also less likely to have errors.
In short, a server is a valuable tool that makes it much easier to spread data, update files, and access that data when needed. You may have a computer acting as a server already!
Crash
That said, a server is ultimately still a computer. It may be a particularly powerful one with a lot of memory and storage space, but it is still a computer! And computers can sometimes crash, sometimes catastrophically!
When an application crashes, it means that it’s encountered a ‘fatal’ error and will exit out rather than continue trying to run. Not all programs do this – some will continue trying to run, eating up all of the device’s memory, and ultimately causing the computer itself to crash when it runs out of resources too, which is why it’s usually better for the app to exit out even if it means losing some data.
When the computer crashes, it’s because the entire operating system has encountered a fatal error. Lack of resources (see ‘a single program is trying to use everything’ above), missing bits, a corrupted file it was meant to have, the list goes on. A lot of things can cause crashes, and while a lot are preventable, a lot are not and have to be monitored for instead.
A server crash by itself is obviously not great. While that server is down, it’s files are inaccessible, which can create a lot of dead time for employees that should have been used generating revenue or working on necessary tasks. Instead, everyone is sitting there and waiting for it to resume function. And, depending on what caused the crash, that might take a while. Worse still, computers can crash so hard that they lose their data. It’s unfortunate – there are several ways the computer tries to prevent it – modern operating systems do it less often – but it can still happen.
So how do you prevent it?
Maintenance
Remotely-controlled and accessed servers will basically always have temperature monitoring. This is because computers stop working as well once they get hot: memory is magnetic, and magnets themselves slowly lose their magnetism as they heat up. It’s why superconductors have to be chilled to nearly absolute zero before they can do those tricks that make frogs float! MRIs, or Magnetic Resonance Imaging machines, need liquid helium chilling the big magnets inside to crank it up enough to make images, on the same principle.
Therefore, servers must not get too hot, or they’ll break and crash. Regular inspections of the cooling equipment can prevent this, and monitoring the temperature of the server can give warning that a cooling instrument has stopped working before the thing gets all the way up to 250 degrees Fahrenheit. Even smaller servers should be regularly dusted or blown with canned air to ensure any cooling fans built in are still running smoothly.
Additionally, power surges can be dangerous for computers! There are multiple parts inside the computer that will stop working if they simply receive a static shock, never mind a full surge. As such, it’s important to keep servers behind surge protectors to avoid loss due to electrical malfunction, although it’s less common than heat death.
Additionally, computers do not like being dropped or hit – servers over a certain size are usually kept on wheels to save people from needing to pick them up to move them, which would then come with the risk of dropping it!
Prevention
Regularly updating a server may seem like an unnecessary time suck, but it’s better than the alternative where the software crashes because it encountered an error that should have been fixed five updates ago. Alongside software, the hardware should also be regularly updated. SSD hard drives are sturdier and less prone to problems than the old standard HDD hard drives are due to having fewer moving parts inside them, so when they become an option for a business, it’s smart to upgrade.
Quality surge protectors don’t cost as much as fixing the server would, and while temperature monitoring may be annoying to set up, it is – once again – less annoying and time consuming than fixing a server that has melted down.
An ounce of prevention can save pounds of cure!
Recovery
So, say something happened to the server – there was no monitoring, and then the server crashed because the temperature inside got so hot that the solders and glues holding the computer together started failing, or the thermal paste melted out. Maybe it got a power surge, maybe it was just unplugged during an update and now won’t reconnect.
The server has crashed hard and it won’t reboot properly. What now?
Unfortunately, the options are limited, and all of them are going to take time. It’s why maintenance is so crucial: there’s no easy answer for a broken-down server.
Hard drive recovery, especially off of a computer that crashed for a reason like excessive heat or hardware malfunction, is not easy, and is very expensive. The price goes up the bigger the drive is and the harder the job is – a drowned hard drive with 5 TB of data is going to be a very pricey fix, should that ever happen. The same goes for surges – at some point, the goal will have to be data recovery and then getting a new server, because a surge can totally toast the parts of a computer that make it a computer.
In essence, hard drive repair and recovery are not magic bullets. They also take a lot of time even when it is possible to recover data, sometimes weeks, which can be real agony if there was any essential data on that hard drive that wasn’t backed up. If the server is needed for employees to do their jobs, remote in, or anything else, they may be left sitting idle for weeks due to a problem that would have been 40$ to prevent, and costing money the whole time they’re doing it.
This is why prevention is so crucial! Regular maintenance can save a lot of pain and money.
For Your Consideration
If server management and monitoring sounds like a serious pain in the butt, or something you simply don’t have time for, consider us! An MSP (managed service provider) is a great way to ensure your IT work is up to snuff and ready to go for the long haul. Whatever path you choose, though, one thing is vitally important: don’t ignore a server!
If this sounds interesting, consider scheduling a call here: https://elixistechnology.com/contact/
