Hi All.

Running BAAN LN 6.1, on Windows Server 2003, Enterprise, R2, with SQL 2005.

Was wondering if there was some recommendations of how I should prepare this system for disaster recovery.

This is my current plan:
Backup Process:
(1) Hourly log backup
(2) Nightly full backup of system state and data

Restore Process:
(1) Load base OS
(2) Partition drives
(3) Restore system state and data
(4) Restore logs close to point of failure.

Our idea is to be close to an hour within when the problem occurred, but I know that is not always the case, but we can at least try, right?

(Also, if someone has documentation of using this within VM, that would be nice, as our original install excluding installation to a virtual machine, which would also simplify backup/restore for us)

Hi,
It would be worth the effort to look into Image Backup. This can eliminate installing everything from scratch. All you got to do is to restore the image backup of the drive (preferrably the system, application and data drives) and then restore the latest db backup (be it full or differential). License may have to be validated again (as was the case with Baan4). The beauty of this is that you could restore the image on a dissimilar hardware in case your Live server is inaccessible.

Tested (read revived a crashed Baan) successfully on Win2k3+SQL2k+Baan4 W/SP19 using Symantec Backup Exec System Restore (SBESR). It works even on a VM Server.

sk

Our idea is to be close to an hour within when the problem occurred, but I know that is not always the case, but we can at least try, right?
You also need to establish a corporate SLA (Service Level Agreement) and also a RTO (Return to Operation) after a Disaster and this can determine the right Solution for the situation.
(1) Hourly log backup
(2) Nightly full backup of system state and data
You need to have enough disk space to handle that.
Your backup process also determines if you might add some activity on the CPU.
Restore Process:
(1) Load base OS
(2) Partition drives
(3) Restore system state and data
(4) Restore logs close to point of failure.
In a real DR event, you rather not waste time to reload the entire OS. You rather have a mirror machine ready to use so your RTO is quick enough. Maybe you can explore the standby database with Dataguard, disk mirroring, live replications or other technologies.

Adding to Sukesh75's point - you should consider a HA (High Availability) for the SLM (license server)

BAAN 6.1 was here before I arrived, and I have no training on it. I think the configuration is sub-optimal [but what do I know?] All the staff involved in the installation [February 2007] are no longer here. If this keeps happening, we will need to outsource IT just to get stability in IT (sad, huh?) [My role is moreso making sure systems are backed up, and that they can be restored, than being directly involved in the workings of BAAN, so anything to make it simple and troublefree is the best idea.]

You also need to establish a corporate SLA (Service Level Agreement) and also a RTO (Return to Operation) after a Disaster and this can determine the right Solution for the situation.

Excellent suggestion. We currently do not have this for any of our systems. (Every day, I feel like my education in IT is backwards, as the further I go, the more I see basic things like this, that I need to educate myself on.)

You need to have enough disk space to handle that.
Your backup process also determines if you might add some activity on the CPU.

With regards to the logs, we have the disk space to handle that. With regards to the CPU load, the transaction logs should have minimal impact, and we do the full backup when the users (should be) asleep.


In a real DR event, you rather not waste time to reload the entire OS. You rather have a mirror machine ready to use so your RTO is quick enough. Maybe you can explore the standby database with Dataguard, disk mirroring, live replications or other technologies.

Ahh, there is my problem. I do not have a mirror machine ready. I have hardware, so maybe this should be my plan:

1. configure identical system
2. implement sql mirroring
3. decide failover criteria
4. test failover

Adding to Sukesh75's point - you should consider a HA (High Availability) for the SLM (license server)

I do not know enough about the configuration of this system to determine what this is exactly.


Hi,
It would be worth the effort to look into Image Backup. This can eliminate installing everything from scratch. All you got to do is to restore the image backup of the drive (preferrably the system, application and data drives) and then restore the latest db backup (be it full or differential). License may have to be validated again (as was the case with Baan4). The beauty of this is that you could restore the image on a dissimilar hardware in case your Live server is inaccessible.

Tested (read revived a crashed Baan) successfully on Win2k3+SQL2k+Baan4 W/SP19 using Symantec Backup Exec System Restore (SBESR). It works even on a VM Server.

sk

OK, I'm getting closer on this, so this is what I think I need:

1. Dupe System & Apps for HA
2. Mirror DB & Data for HA
3. HA for License Server
4. Full System Snapshots for Backup
5. Quarterly Testing

OK, Just to be clear, these are the components of BAAN LN?
1. Apps
2. Data
3. DB
4. License

Hah, funny, the list reads like some other applications that I already have, it looks more simple this way ... thanks!

I apologize for my ignorance. Thanks for your help. I am supposed to be configuring this in August-September (anything is better than relying on tape).

I was told by the BAAN guy who was here for the original installation before I came on board [he's no longer here] that the BAAN LN software is not a clickety-clickety installation, and I am hoping I can get it loaded onto the mirror system from the backup.

I'll let you know how well it goes, and feel free to post more tips in the meantime.