Migrating to a data center facility does not necessarily resolve issues or automatically alleviate the headaches associated with server administration. Opting for Managed Services is one of the most important decisions an IT manager can make; it can help focus on the major issues and relieves the problems of data loss and outages.
IT managers are often overwhelmed, not only by day-to-day maintenance , but also by major applications for which they are responsible. This could disrupt necessary upgrades, such as delaying upgrades on physical hardware, which could result in continuing to operate very old servers. It is recommended that your servers be replaced every four years. However, life cycle depreciation varies from organization to organization. Ideally, you should ensure that your servers are replaced and updated according to a predefined matrix and schedule. Life cycle depreciation can be eliminated by leasing servers for a set period of time; new hardware upgrades can then be requested upon termination of the lease and contract renewals.
Computer equipment is often located in back closets without adequate cooling; this poses even a larger threat to your servers when the air conditioning unit is turned off during weekends. A sufficiently ventilated, cool, and static-free environment is essential to the longevity of your servers and your business.
Reduce downtime and site outages by ensuring your traffic load is properly balanced. Nothing spells disaster like having a server or site go down due to unexpected traffic volumes! Ensure your load is split between hardware located in one data center facility; you also can spread it throughout the nation or even globally. Geographically balancing your traffic load can help minimize pesky latency caused by your products’ global domination!
Maintenance, security and common sense will also help to reduce downtime and outages. Properly securing your information behind firewalls and performing routine maintenance on your equipment will help reduce your risks of costly security breaches that could bring your network down. System hardware maintenance is also crucial. For instance, making sure your RAID arrays are working properly will help eliminate the chances of experiencing downtime. Typically when a RAID is reorganizing your data, you are twice as likely to have another drive fail – resulting in a server outage due to stress on an already ailing drive. Another example is to ensure security patches are updated on a regular basis to alleviate the risk of suffering from an outage or downtime.
Keep technical contacts updated with your service provider. Your service provider will perform routine maintenance on its systems from time to time. Standard procedure is to notify all customers; however, if your CTO or ancillary technical staff members are unaware of these maintenance windows, it could be perceived as an outage.
Proper notification will ensure that you have a plan in place to ensure proactive steps are taken to prevent being affected by system maintenance and reduce downtime.
Disaster recovery planning is crucial. Ensuring you have a Business Continuity and Disaster Recovery plan in place is one way to prevent an outage. Having one in place, however, won’t prevent an outage and not knowing how to implement it won’t mitigate loss associated with downtime. Make sure your staff is properly trained on the policies and procedures in order to facilitate a smooth transition to help ensure your engineers meet their recovery time objectives.
Ensure you have redundant carrier connections. If purchasing connectivity from your data center provider, make sure it has a blended network composed of minimum of two carriers and use Border Gateway Protocol (BGP). BGP is used to connect ISPs to each other and/or to connect different physical sites that are owned by the same company. It is the language routers use to speak to each other. Border Gateway Protocol helps to ensure you are receiving the fastest connection possible. Always opt for a redundant backup connection even if you already have a multi-homed BGP connection. This is like a small insurance policy that costs a little extra but can pay off big.
Even the world’s most advanced, well-planned data center operations are not immune to outages due to the unexpected. According to a recent study by the Uptime Institute, “human errors account for nearly 70% of reported outages.” Data center facilities are run by people, so be prepared by ensuring you have a Business Continuity plan in place.
Cold Site storage of your data backups , and even a clone of your online web presence is always a good idea to avoid costly repercussions of downtime from an outage caused by human error.