Hal Kaplan <> wrote: > => -----Original Message----- > => From: [EMAIL PROTECTED] > => [mailto:[EMAIL PROTECTED] On Behalf Of Jim Felton => Sent: > Sunday, October 08, 2006 08:54 => To: ProFox Email List => Subject: > RE: [NF] - SQL- Database Replication => => Stephen, The two server > are mounted in the same rack. They => are used for a broadcast TV > automation system. The SQL => servers hold the Metadata for video > recording and playback. => The Metadata comes from user input, IP > based Internet and => satellites feeds, Baseband recordings from tape > or live => broadcasts, automated video ingestion and automation > => control. We aren't looked for two systems updating one > => database at the same time; we are looking for real time > => redundancy to protect use from hardware failure. We are => trying > to eliminate a singe point of failure and these => servers are the > last piece to the puzzle. => > > All right, what am I missing here. IMHO, for such a critical system > mounting both in the same rack is one of the dumbest things you can > do. Am I wrong? Please explain. TIA
There are two methods of creating a seamless uptime for your data. You can create a cluster where multiple SQL engines are waiting for the cluster to tell them who is running. Those SQL servers are both pointing to the same database on a SHARED data device on your network. You can also turn on replication and have the SQL server dispatch data to a different server anywhere in your organization. Replication demands some switch between the servers when one goes down and the other picks up. You also have to determine how up to date your going to make these separated systems. Do you use snapshot, merge or transactional methods for "replication" ;-> They each have their own + - lists. There are quite a few vendors who are in this space to automate this whole mess. Price tags are large, and maintenance fees for them are almost as profound. You will be paying on the # of servers, # CPUs, that are part of your environment. Clustering does that for you with no loss of data, you do loose uncommitted data when the switch takes place. Clustering is free from M$, well as free as that second server, OS, SQL license get. But you do have to pay that same price for EVERY method your going to look at. Now Q about having both primary and backup on the same rack? I know that in my last gig I had that set up and did a logshipping through a crappy VPN back to my business office. Distance was only 10 miles, but bulk transactions failed in the every 15 min push because data size didn't fully copy over. So watch out on those little aspects of covering your assets. Stephen Russell DBA / .Net Developer Memphis TN 38115 901.246-0159 "Our scientific power has outrun our spiritual power. We have guided missiles and misguided men." Dr. Martin Luther King Jr. http://spaces.msn.com/members/srussell/ -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.407 / Virus Database: 268.13.1/466 - Release Date: 10/7/2006 _______________________________________________ Post Messages to: ProFox@leafe.com Subscription Maintenance: http://leafe.com/mailman/listinfo/profox OT-free version of this list: http://leafe.com/mailman/listinfo/profoxtech ** All postings, unless explicitly stated otherwise, are the opinions of the author, and do not constitute legal or medical advice. This statement is added to the messages for those lawyers who are too stupid to see the obvious.
