What is the real cost of planned downtime? Does it really cost your company anything at all? (Some have suggested that productivity might actually increase when e-mail is down!) Does reducing planned downtime really justify the added cost, a very significant cost, of clustering, especially considering that planned downtime can be taken at times when few users are on the system? It seems to me that clusters for Exchange seldom can be cost-justified.
Ed Crowley MCSE+I MVP Tech Consultant Compaq Computer "There are seldom good technological solutions to behavioral problems." -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Exchange Sent: Thursday, January 17, 2002 9:09 PM To: Exchange Discussions Subject: RE: Clustering Exchange Hmmm...So everyone likes to cut down clustering. I don't fully agree. I have worked with it tons, both on the 5.5 and 2000 platforms. It is more complex, more things can go wrong, many times 'the clustering part of it' lags behind the other parts of the program, third party products aren't always cluster aware and therefore can cause problems, human error is much more common because people aren't sufficiently trained on it, etc. But, if you educate yourself on the technology, work with 3rd parties that do consider clustering, etc. it is possible to, first and foremost, have less planned downtime. That really is one of clustering's major benefits. You can deploy service packs, etc. on the passive nodes while the active one is still running, and effectively cut a 1/2 hour downtime situation to 2 minutes (if everything goes well ;-) ). Also, if the hardware, OS, or a service, goes South on a system, failovers happen quite gracefully (given you're up to date on service packs) and you will have the service back up faster than if you weren't clustered. I've had systems with 5,000 users on them failover with minimal reports to the customers help desk...it does work. If you're in an environment that has people who know what they're doing, and the decision makers above are willing to spend the money for adding possibly another 9, clustering can help. If you are new to the technology, let others do it - your stand alone server will run just fine and will be easier for you to maintain. If you do go with clustering: Don't do active/active clusters, and don't forget that the single point of failure is your SAN/external disks - clustering won't save you from database corruption or external disk failure. In your situation it certainly sounds like not clustering is the right thing to do. I just wanted to defend the technology a bit, because I feel given the right circumstances, it performs as advertised. 'Hope it helps, Per Farny Senior Network Architect Goliath Networks Inc [EMAIL PROTECTED] -----Original Message----- From: Callan, Chris [mailto:[EMAIL PROTECTED]] Posted At: Thursday, January 17, 2002 11:34 AM Posted To: Exchange Conversation: Clustering Exchange Subject: Clustering Exchange My immediate supervisor mentioned that when we finally get new Exchange Servers that we should have them clustered. Now I have never clustered servers before and wouldn't know how to start, but I just wanted to get everyone's opinions on the subject to begin with. How hard is it to do, and how is it to maintain. What are the pro's and con's. Any help would be appreciated. __________________________________________________ Do You Yahoo!? Send FREE video emails in Yahoo! Mail! http://promo.yahoo.com/videomail/ _________________________________________________________________ List posting FAQ: http://www.swinc.com/resource/exch_faq.htm Archives: http://www.swynk.com/sitesearch/search.asp To unsubscribe: mailto:[EMAIL PROTECTED] Exchange List admin: [EMAIL PROTECTED]
