Four of the Most Troublesome Server Problems

                As any large corporation can ensure you, the longer a server is down the more profit is lost. For large corporations which lean heavily on their IT infrastructure, if a primary server array falls off the grid they stand to lose hundreds of thousands of dollars in lost sales, employee downtime, and resulting stock drops. While you may not be at that level with your current server configuration, there certainly is something to be said about protecting yourself from server failure. Server problems will manifest larger problems within your business unless they are handled quickly and concisely. Don’t know what you ought to be looking out for? No problem. Here are four of the most troublesome server problems, along with a few helpful hints regarding each.

Hard Disc Usage

Your hard drive is among the most important components within your server. All of your important business data, from invoices, product locations, client data, employee data, backups, etc. will be stored on your server’s hard drive. Consequently, backups are always a good idea. But even if you do have your data backed up regularly on a secondary drive, unless you have a redundant option a drive failure will cause a considerable amount of downtime. Look out for sudden increases in logs, warnings that your system is running out of storage (especially if the graded capacity isn’t reached), unexpected data loss or directory errors, or even unexpected noises.

Memory

Your server’s RAM is what provides the processor with ample “working space” making for speedy and efficient multitasking. System memory is especially important regarding servers, as these machines are tasked with a great deal of operations at once. Memory is most often replaced rather than fixed. If you notice that your processes are bogging down or running excessively, or if the CPU usage spikes suddenly, it’s likely that you may have a RAM failure.

Server Cooling System

Your server ought to have a temperature gauge, either a physical model which monitors the core temperature or an electronic thermometer with a peripheral display. If you notice that your server’s temperature is exceeding working bounds, chances are your fan or liquid cooling system is failing. As the internal temperature rises, you run the risk of frying your motherboard or damaging the components of your processor and memory. Oftentimes, this can be solved by performing physical maintenance on the fan or by adding a second fan to the assembly.

Power Supplies

The power supply is perhaps the most fundamental component of your server. Unfortunately, power supplies often fail without warning. Always check to make sure that all connections are secured before assuming the power supply has failed. Many people opt to instead have redundant power supplies. I keep an extra 300-1848 power supply on hand should my current supply fail. This effectively minimizes potential down time due to a server failure.

Computer Upgrade: Orders of Priorities

                The newest and best computer is right around the corner, and with every new release of an updated model, your current computer becomes just a little more dated. Whether it’s a personal desktop computer, an ultraportable laptop, or a high-powered server, sooner or later it’s going to be phased out. That’s unfortunate, considering that all of these computers were and remain major investments. So what do you do when you need to boost your computer’s performance to keep pace with the rest of the industry standards? If you’ve got a budget in mind, you may not be able to do a complete renovation on your computer, so we’re not talking a brand new configuration inside an old case. You may only be able to afford a single upgrade for your computer. So the question remains: which computer component warrants the investment? Here’s a short list arranged by priority, with an explanation as to why.

First And Foremost – The Processor

If you’re intending to upgrade a laptop, this one can be a bit of trouble. A lot of laptop boards integrate their processor with the board or have very restricted socket types. It’s easier when it comes to desktops and server computers. The processor will show the most flat out improvement in power and efficiency. Multi-core processors or multiple internal processors will allow your computer to handle more in a shorter time. Look for a processor with a high clock speed, and ideally, the ability to overclock. If your server is running a lot of processes, multiple cores are especially important.

Second – Storage

Some people might think it strange that storage comes in second. After all, it’s just a matter of capacity, right? That’s part of it, but not nearly the whole story. Upgrading your storage can result in increases in performance, provided that you make the right upgrade. Look for a hard drive which has a higher spindle speed than your current drive. This offers faster data writes and retrievals, so anything stored on the drive offers faster access. If you’re looking for reliability and speed, consider upgrading from a traditional magnetic hard drive with moving components to a solid state drive, which currently offers the industry’s fastest performance.

Third – The Memory Expansion

For personal computing, RAM is often heralded as one of the most important upgrades. For most people, you can scoot by with four to eight gigabytes of memory. Memory is especially important if you intend to do a lot of multi-tasking. If you only run three or four applications at a time, you won’t see a great increase in performance from adding RAM, as most applications don’t eat up anywhere near a gigabyte of random access memory. However, in server applications, a lot of RAM is very important – one of the reasons why my number one priority in upgrading my server was the addition of Sun Microsystems 8GB servers memory upgrade. As servers are tasked with multiple processes at once, sometimes number in the several dozens, the additional RAM does produce a noticeable increase in performance.

Where to Sell Used Servers

So you have finally made the hard decision to say goodbye to your old server. It’s probably been collecting dust in some supply closet for a few months, and you just can’t seem to find a use for it. Your new server is working just fine, and you’ve already transferred all the data you need and wiped the drives on the old server. So what do you do with that outdated piece of technology? If you’re considering just taking it into the recycler…stop! That old server still has value, and while you may deem it without value, there is plenty of money tied up the system. Getting your money’s worth out of that old server is one of the best ways to get a return on your initial investment. Selling a used server is a great way to recoup some of that cost. But is there even a market for used servers? How would one go about selling his or her server? The short answer is yes, and there are two ways to go about selling your server: person to person and direct sale.

Person to Person

As this method suggests, person to person sales are carried out among individuals across a variety of mediums. The most common solution for person to person sales necessitates some sort of third party facilitation. The third party might be a newspaper providing you space for a classified ad, or it might be an ecommerce site allowing you to post. Ecommerce is a viable option for selling your used server, but beware: as you enter the ecommerce arena, you will be competing for a low price against others in much the same situation as you. If you have the lowest price on a site like Amazon or Ebay, chances are you will be able to sell your server. However, if you’re holding out for the right price you may be waiting for a while.

Direct Sale

A direct sale involves approaching a used hardware or computer business and negotiating a flat price for the exchange of the server. There is no “bidding” in the sense of ecommerce auction style purchasing, but you may be able to negotiate a higher price, depending on the quality of your server. When selling your server directly, it is important to get a quote and verify it at another buyer. Whereas you were competing for the lowest price online, you will now have a little bit of leverage as potential buyers quote up against one another. Even if your server has no value, many computer businesses will be able to provide you with a scrap value for the server. With the precious metals contained within the server, you will still be able to garner some small return from the server.

Bottom Line

Direct sale is preferable to person-to-person sale, and often provides you a more realistic return on your server. The ability to request a quote and verify a quote before you sell used servers greatly increases your chances of getting the best price, whereas online selling will cut your margin of return further down.

Expanding Server Potential

All too often, technology is quickly outpaced by the some “newer, better, faster” replacement which is, in reality, only marginally improved over the previous model. It happens all the time with smart phones, as an example, but who really needs two cameras on a phone or a female HAL artificial intelligence? Over time, however, these small improvements add up to the point where one model is truly outpaced by a newer model. But with major investments, one can hardly afford to get the newest and fastest model every time a substantial improvement is made. When it comes to servers, the best way to plan for future needs while maximizing on the use of an older server is by planning for scalability.

Individual Scalability

The first aspect of scalability is that which applies to individual servers. When choosing a server for your business, you should always plan for future upgrades. Scalability comes in as we consider the question To what degree is this server scalable? If the server has two gigabytes of installed RAM but only allows for a total of four gigabytes, that server may not be the best investment. If a server allows for up to sixteen gigabytes of expanded memory, that is the much better choice. Individual servers with greater degrees of scalability will allow for the easy installation and removal of processors, memory, and hard drives, depending on your individual needs.

Multiple Scalability

When you are looking at expandability for multiple server configurations, you will want to look at the scalability for multiple units. Tower servers are generally good for small businesses and operations. However, multiple tower servers wouldn’t be optimized for larger businesses. Tower servers are generally standalone units. The best choices for larger-scale deployments are rack servers and blade servers. Rack servers are those which are self-contained units but which are placed in rack mounts. These are the most common choice for scalability; as multiple servers are required, each can be individually scaled, and the entire compliment of servers can be scaled as a whole. Blade servers consist of one large blade chassis which supplies both power and cooling to the individual blades. The “blades” have all the components of the actual server, and are a great choice for scalability as you can engage only a single blade or several depending on your business needs.

Within rack and blade mounting, individual scalability holds true. However, planning for future expansion demands you be mindful of how server deployments will work. Both racks and blade chassis configurations have a limit to the servers they can support. However, expansion options are available. The IBM I-series, for example, allows for expansion drawers for additional scalability. With the addition of a 5796-8203 expansion drawer, the potential of the IBM system is expanded beyond its maximum individual configuration. If you don’t want your server(s) to be passed up and rendered useless, you must plan to upgrade individual servers. Additionally, you must allow for multiple server expansions. Allowing for such growth will ensure your investment yields the greatest benefits.

Not a Fan of Server Meltdowns

Accidents are a part of life. They’re unavoidable, but oftentimes we end up just kicking ourselves, saying “If only I’d have seen it coming!” There would be far fewer cars in the ditch, skinned knees, and burnt dinners. Luckily, the majority of unfortunate incidents lead to only minor consequences. When it comes to technology, the majority of “accidents” we have turn out to be nothing more than a file we didn’t save soon enough or an email with a typo. However, every once in a while, technology takes a turn for the worse. Oftentimes, the culprit behind these accidents is something we neglect up until we have a problem with it. Your server is a major investment, and protecting it from harm should be a primary concern. So what’s the specific culprit I’m talking about? It’s your server’s fan. So take heed, and take care of your server fan before it quits on you!

Why the Fan?

It seems like such an innocuous component in your server, right? A fan, while not the most noticeable or glamorous parts of a server, is still absolutely essential to the proper functioning of your technology. Fan problems can lead to devastating results if not taken care of quickly and effectively. If your system isn’t cooling well enough, your processor and motherboard can be damaged to the point where they require replacement. That will be an expensive undertaking, one which would be best to avoid. Here are three signs of impending fan failure, which, if noticed, should be taken very seriously.

Excessive Noise

While a normal server fan will run a little bit loud (it’s doing a lot of work to assist in the removal of that heat, after all), if your fan is running excessively loud or making odd whirring, clanking, or scraping noises, it’s time to call somebody in and shut down your system for a quick repair. Fan repair/replacement is relatively inexpensive, and should take no more than twenty minutes if done by an experienced professional.

Internal Temperature

This is one of the best ways to anticipate a problem with your server fan. Most servers will throw up a warning if the internal temperature climbs too far outside of normal operational parameters. Many servers will automatically shut down, as well, if the problem goes without service. However, why wait for it to become a bigger problem? Instead, monitor your server’s internal temperature from time to time to make sure it’s not getting too hot. If the environmental conditions are stable in your server room but a system’s core temperature keeps climbing, chances are the heat exchange is failing.

If you do need to remove a fan, you will need to remove the entire tray along with it. We always clean our fan along with its fan tray assembly. Your fan tray will collect dust and various particles. We clean our 541-0645 assembly once a month. With proper maintenance and careful monitoring, you need not risk the possible damages of a server fan failure.