The time is 3:00 AM. You are fast asleep – then the phone rings. “Hello?” you say. “Server is down,” you hear through the tiny (and tinny) cellular phone speaker. This means one of two things: You will either have a very bad morning in a few hours when you have to go fix it, or you are going to have a very bad morning starting immediately when you have to go fix it (more often the latter, of course.) Why must we suffer so?
Well, for several years we haven’t had to, but not everyone realizes this yet. For a relatively cheap investment that easily covers its return after the first server-down scenario, a remote access card can be included with your server that allows for offline-servicing and direct console and KVM (keyboard/video/mouse) connection to a server that is either in an offline or “hung” state. This offers several advantages in the world of IT. An entire infrastructure can be deployed by an expert from anywhere in the globe starting from the ground-up with the operating system install. A server that needs you to press the “F1” key during a reboot because of a failed component can be brought online by the technician on-call before the official start of the business day even draws near.
Not sold yet? Imagine what the means if you outsource your IT support – response times can be greatly reduced because blocks of time do not need to be set aside for the dispatch. Travel & boarding costs are reduced tremendously or eliminated entirely.
Look for integrated baseboard management and lights-out cards from server manufacturers such as HP, Dell and IBM.
- by Andrew, "My Goatee Does Not Make Me The Evil Version Of Myself", Levin
Check out this behemoth.
It's currently the world’s fastest super-computer according to top500.org, a site which publishes a list of the fastest and most powerful machines every six months or so. The Tianhe-1A is owned by China who has just recently surpassed the US as having the fastest super-computer. Although all of its ~14,000 Intel processors with 186,368 cores came from the United States, it’s inter-connect was designed and built by the Chinese. The inter-connect is a core component that allows all the processors to connect to one another.
Back in 2002 the Japanese developed a super-computer which surpassed the US who held the number one ranking at the time. Shortly after the announcement though, the US receiving funding from various organizations and was able to reclaim the number one spot with the Cray XT5-HE. So we will see what happens now with China standing tall on the forefront.
So what is the need for such large scale and powerful machines? Scientific purposes is the obvious answer that comes to mind, which is indeed pretty much the main use. There are various applications of this however, such as quantum mechanics, military research, weather modeling, atomic calculations, etc. I want one.
- by David, "And Don't Call Me Shirley", Quiram
Have you documented your disaster recovery plan? I am not talking about listing some tasks you want to do or where you kept your backup cd, but a full planned out process to get the IT portion of your company running again in a timely and efficient manner?
Most companies have not and some will not. Having gone through a disaster for a company which did not have a disaster recovery, I have seen where the value of having a process laid out will make the recovery less stressful for you, the owners, and the staff.
Having the plan figured out before a crisis gives you several advantages:
It gives a level of confidence in the systems, your ability, and the future in case something happens.
It begins the exercise of how would I do this before things go wrong, as opposed to how do I do this with things going wrong.
It gives you a framework to work with. No static plan will ever fully cover a dynamic event. The saying is the first casualty of any battle is the plan. But having steps mapped out, priorities determined, resources identified, and an understanding of what needs to happen will allow you to be more maneuverable to the changes in the crisis.
By documenting your plan or process for recovering your IT resources you will be showing the owners that you are thinking of the future. It isn't doom and gloom thinking, the reality is that things go wrong, thing will break, and it is never convenient. Plan ahead to reduce the issues and risks you will encounter. If you have a well thought out and tested plan in place, you will be able to support the company’s Business Continuity plan. Remember the BC plan is different from the DR plan. The DR plan is there to support the BC and should be integrated into it. The DR provides the tools for the Business to move forward after a disaster, but we need to make sure we have the tools ready and available.
I have written up several disaster recovery plans for companies that range from complicated multi server/multi-site to single server at someone's home office. Each time the planning and documentation process has proven to be invaluable. It is a time for discovery, identifying issues that would hamper the recovery process, and gaining a better understanding of what needs to happen when. The plan will also get the owner's buy-in to the process beforehand and set up realistic expectations BEFORE the disaster happens. This is very important. If you can show that the recovery effort will take 24 hours for the core infrastructure to be re-established, it will give you the time and room to do your job without the owner and management worrying what you are doing.
The plan provides the framework to address the crisis. No matter how well I have written a plan, there are unexpected events that occur which can derail the plan. Even in testing things happen, replacement servers fail, key equipment doesn’t show up, staff members are sick, network devices power supplies fail. All these have happened just in testing. I welcome these issues in testing because it adds another stress test to the plan and staff working with the plan. By identifying the priorities in the recovery, you can react and re-task resources based on that. The advantage of having an establish priority in recovery is that if you are questioned on the reasoning you have a document which was agreed upon beforehand by management and owners that you are following.
The planning of a DR is only the first step in the whole process. But it is a step that should not be taken lightly. Understand that once you start you are going to find issues with in the systems you are backing up, the thinking and views of management, owners, and staff. By getting all of this address and the plan laid out ahead of time, you will save time, reduce frustration, reduce risk, and increase success in the event of a disaster.
Make sure to take the next step and test the plan through a walk through and then an actual recovery with test machine to simulate the disaster. This will all be time and money well spent when the disaster occurs. If you'd like to have Trigon review, or create a disaster recovery plan for your business, be sure to contact us.