Data Backup and Recovery

Heartbeat (Linux HA) and NetApp?


Does anyone have any experience setting up a high  availability two node Linux cluster using heartbeat ( and  NetApp storage (preferably using SnapDrive for Linux)?

Basically I would like to mount the same NetApp LUN over Fibre  Channel to two servers in an Active/Passive mode (only one server can  access the LUN at a time)

If so, is there a How-To document or at least a list of things to watch out for?




For DataFabric Manager, we used Veritas to cluster 2 Linux hosts.

I used SnapDrive to create the LUN from one host, & then from another host issued a connect command.

From my experience I can list a few of the points to take care of

Basically if an application needs to be accessing the shared disk, then

1. Disable automatic startup of application during reboot on both the hosts

2. At any point only one host must be accessing the shared disk, so that data is not getting corrupted

You set the cluster server software to mount the shared disk on only one host at a time.

3. And before the application service is started, the DiskGroup & logical Volume must be enabled & disk mounted.

You can create a dependency tree, using cluster software.