Introduction to the planning of a data center cold shutdown
An important goal of data center, it is a small server room in your basement or in a large server farm with thousands of Severs, is a continuous service to network services and information technology they support. However, this is not shut down without a sufficient amount of maintenance from time to time, even feasible, sometimes cold.
Although theoretically a data centers are expected to keep their networks running continuously, they must make room for scheduled maintenance windows. Sometimes a cold shutdown, may be necessary. So if a cold shutdown is necessary?
1.dat Move Center: If your data center is moved from location A to B, then most likely you will need to provide a window of time to stop to change the things A and to move as are rare, as they often service disruption, so that it can not tolerable in the area. Such initiatives are often carried out in small pieces a few servers at once.
2.Data center expansion / growth: From time to time you will need themselves to demand the center of the other data, the additional wiring or construction can grow requires a cold shutdown for safety reasons.
3.Original construction or damaged cables, if the original design of the computer center was inadequate or inaccurate, you may need It Down Cold closed and set design before taking it up with. One example is a computer center, the backup battery UPS delivers only a part of the site data center. For example, in a recent case one of our customers, the VoIP equipment in a closet was not an emergency generator to him. Thus, in the case of failure of power failure, phone lines and voice mail systems would. To remedy this situation, the electrical work which requires a necessary stop data center for reasons of safety for electricians.
Cold closures are not easy. The shorter the window, and tolerance for downtime, the more the project. A weekend is usually less time you need to get a cold data on 50 servers in the area and bring it back to be fully functional.
Here are the reasons. On shutdown of a computer center with 50 servers, for example, there are chances that at least one or two servers may have been old, and perhaps the last round. When she returned to a halt, they can not work as expected. This requires to activate the server or device owner or build a system backup or restore / backup. This work can often take some time depending on the complexity and sensitivity of the systems.
arrest is usually from top to bottom and start up. What this means is that the crash occur in the following order:
• presentation layer servers and Web servers
• Application Server, the evaluation of the virtual server
• database servers
• Storage and backup servers
load balancers and other devices •
• Communication equipment such as server, telephone, fax server, etc..
• Network devices such as routers, switches and access servers
• Low network cabling and infrastructure equipment
You are back in reverse order.
In planning for computer maintenance
Please note it is important that such an approach is risky and often with significant loads in the event of a prolonged stoppage levels mentioned above. It is often useful to call a list of technicians who can help you if parts of the closure does not hold as planned. Note also that critical documentation may be printed and kept ready locally on a computer, because most documents in systems that are affected due to the closure could be planned. If you are new to the maintenance of the data center, you can hire or consult with experts to ensure that everything runs smoothly.
Moving a data center