High Availability and Backup options... what's everyone using

[mlchris2]

We are currently running NeverFail for Exchange/File Server/SQL Server on our production DB servers as our High availability solution...

We currently use a Web-Based backup solution called eSureIT to backup all our our production servers.

I've been tasked with a project to find out if NeverFail is still the way to go for High Availability or if there is another producted out there that does both High Availibility and Backups or is the solution just to implement Backup Exec or something similar.

I would like to find out if you guys are using any other High Availiblity / Backup solutions? Do you use 2 seperate applications or do you use an application that does both? Anyone using VMWare or the similar?

it would be great to hear what you all might be using and the pro's/con's....


Mark C.

 

[kmcferrin]

VMWare and Virtual Infrastructure is what we use. I'm not familiar with the applications that you are using, but the quick story on VMWare:

You get some blade servers, or a couple of 4-way dual/quad core servers, and attach them to a SAN via redundant connections over redundant SAN switches. You set the servers to boot from SAN and load VMWare ESX. From there, you run everything virtualized and store the datafiles on the SAN. If you need high availability you use Virtual Infrastructure's HA tools and VMotion to allow the machines to be automatically migrated across ESX servers in the event of a hardware failure. With that your max downtime from a server hardware failure is a couple of minutes. If that's too long for you, run your application in clustered mode (SQL or Exchange or whatever) to cover that gap.

For backups, VMWare has a nice backup tool that will back up your VM files. You can also use SAN replication to a remote SAN, snap copies, or whatever other SAN tools you want to make redundant copies offsite.

If you do it right you can run a remote hot site on ESX as well, and replicate your VM files to that site as a backup and DR strategy. Then if your primary site goes down you can simply boot your servers from the hot site and be ready to go.

 

[mrdenny]

We use VMWare for our application / print servers.

Our SQL Servers are clustered for HA.

Our Exchange servers will be clustered soon for HA.

For the SQL Servers DR is handled via Log Shipping from CA to TX for the SQL 2000 servers and via Database Missoring for the SQL 2005 servers.

DR for the Exchange servers is not required by the business short of a tape restore several days after the fact.

Denny
MCSA (2003) / MCDBA (SQL 2000)
MCTS (SQL 2005 / Microsoft Windows SharePoint Services 3.0: Configuration / Microsoft Office SharePoint Server 2007: Configuration)
MCITP Database Administrator (SQL 2005) / Database Developer (SQL 2005)

--Anything is possible. All it takes is a little research. (Me)

http://www.mrdenny.com