Planning a Syncronized Failover 2003 Svr

Posted on 2004-10-05
Medium Priority
Last Modified: 2013-11-15
We're in the planning/testing process of migrating from NT4/Exch5.5 to SBS 2003 (Win 2003, Exch 2003). We're a small non-profit, and have under 15 users doing email and file/print on one server. Our 5.5 db's total 1.8GB in size, and there's not any unusually large volume of email. In the wake of Florida's 'Fantastic Four' Hurricanes, I've been given a small budget to develop some form of 'warm stand-by' environment that could be up within a few hours that would have a fairly recent version of the production environment. We'll be locating this server at a partner's office (hopefully) out of state. I'm not worried so much about the file and print, but more about the exchange side of things.

With that aside, I'm soliciting input on the best way to keep this warm stand-by DR machine synchronized with data no more than a day old. I sincerely would like to avoid any 3rd-party software purchases, but if one could give us on-line failover, and robust fail-back, for less than USD$250, I'm all ears. I've looked at some briefly, but found them to be 800 on up. Here's what I have so far:
Two Small Business Server 2003 DC's, connected via a 256k VPN
Both using DFS for file synchronization
To initiate a failover: The public DNS/MX records transfer, the clients VPN to the DR server, the clients change their outlook settings to point to the other server.
To initiate a failback: Reverse of above, FULL NTBackup *FROM* DR *TO* Master environment, OR possibly using Recovery Storage Groups to get new messages merged back into the master store, along with a datestamp-filtered xcopy to get changed files back.

These are the possible ways I've heard of to do this:
1) Elaborate WAN-SAN Configuration, with Enterprise Clustered configuration ($$$$$, not an option)
2) 3rd-party Tools ($$$, prefer to avoid)
3) FULL Daily NTBackup to disk, which is replicated across the VPN, and restored
4) Full & Daily Incremental NTBackup, similar to above
5) Periodic Full NTBackup, together with Replication of Logs (similar to SQL log shipping)
6) Some on-line replication native to Exchange. I hear this doesn't exist

Well, am I delusional, was this good only for a laugh? Or is this possible? I'm leaning towards 4 or 5, hoping for 6, am keeping 2 in my back pocket. I'm offering 500 points not because it's urgent, or extremely diffcult to someone who would know, but because I'd appreciate some thought-out ramifications of any proposed solutions. Be pragmatic, and offer your experience using the techniques. Thanks much.
Question by:ngray77
  • 3
  • 2
LVL 104

Accepted Solution

Sembee earned 1500 total points
ID: 12229766
6 doesn't exist. It is probably the most requested feature of Exchange that doesn't exist.
Therefore you are stuck with backups. You haven't said how large the stores are, but that link is probably not fast enough to do a live backup over the VPN. You will have to backup locally, then once it has finished copy that file across.
5. Might work. Except you cannot touch the log files while Exchange is running.
Your best bet might be to look at one of the offsite storage companies who collect the media from you on a regular basis and then store it.


Author Comment

ID: 12238797
6) That's what I understand so far.
5) As I said, it's 1.8GB. I don't have a huge problem with stopping exchange services for the 5 minutes in the middle of the night it would take to copy a few 5mb log files to the local outbound synchronization queue, at which point the services could be re-started.

The problem with Iron Mountain and their more local ilk is that it doesn't get me a new server up that fast. To have a DR plan for either scenario, I need to have standby hardware somewhere geographically separate from our location. Who will order the tapes? Who will restore them? How long has it been since the restore was tested? The advantage of shipping incremental data (from a 3rd party, or incremental backups or replaying log files) is that the dat is ALWAYS there, ALWAYS restored, and and ALWAYS tested. Some simple scripting work can alert me of failures, and some more scripting will get me periodic 'full' backups stored on the same DR server. Given the plan so far, the tape starts to look unecessary alltogether. (Always have yesterday's data, Always have a few snapshots of the past few weeks)

I KNOW somebody out there has a similar mandate... anybody?
LVL 104

Expert Comment

ID: 12243006
The solution I am starting to look at is Physical to Virtual technology. VMWARE are working on it, and there are a couple of the third parties doing so as well. Using P2V means it doesn't matter what the host hardware is, as long as it runs VMWARE you can get the network back up and running - I estimate in less than an hour. If it is that key perhaps that is something to look at?

Unfortuantely being a Brit the most I have to worry about is a lot of rain, or a client's converted barn burning down. In that case I the client will have more pressing things on their mind and my quoted 48 hours to get the site back up and running is adequate.
Perhaps you are trying to do too much - has the company actually quantified the loss of server etc and put a value on it? If it is that critical then you can ask for a higher budget and consider other solutions.


Author Comment

ID: 12423984

I've looked into the virtualization route, and it does look very positive, especially in situations where you'd like to recover multiple servers onto one recovery machine. It might run slower, but it's available. I've looked at MS Virtual Server 2003, VMWare, and one other the name of which eludes me.

The real crux is in synchronizing them, which is no simpler with virtualized hardware. In fact, it may be more difficult, because I know of no way to do incremental environment synchronization, just the WHOLE environment. Huge.

Our licensing is mostly donated, I already have another copy of SBS2003 and CAL's. And I can use older hardware, just a few hundred there for an older ebay server. There are a few reasons I'm going this route, including 1) It's mostly unattended. In a disaster, I'm going to tend to my house first. That puts the office back a few days. 2) It obviates tape backups. 3) It's less expensive than 3rd-party recovery services, by far.


Author Comment

ID: 14322725
Forgot this was here. Never got a great anwer, but the first was the more useful.

Featured Post

Get your problem seen by more experts

Be seen. Boost your question’s priority for more expert views and faster solutions

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Office 365 has multiple features and services which are specially designed to help businesses to reduce their dependence on on-premises IT resources. It also offers great flexibility and enhanced security. But like any other data, Office 365 mailbo…
If there is anything erroneous with Exchange Database, it causes a significant effect on email communication till the user remounts the database. Further, database crash directly affects Outlook users due to which they are unable to access their ema…
In this Micro Video tutorial you will learn the basics about Database Availability Groups and How to configure one using a live Exchange Server Environment. The video tutorial explains the basics of the Exchange server Database Availability grou…
Exchange organizations may use the Journaling Agent of the Transport Service to archive messages going through Exchange. However, if the Transport Service is integrated with some email content management application (such as an antispam), the admini…

624 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question