My HACMP skills are 10 years out of date but I'll soon be configuring a 2 site 3 node cluster. Site A will have 2 nodes with mutual takeover and non concurrent vg's and shared storage via a basic SAN and DS5020. Site B will have 1 node and is for DR, it will have its own directly connected DS5020. I'll be using GLVM over IP in Asynch mode. I'm currently plowing through the on-line red books etc but where to begin ? Anyone like to propose a very high level order of work ?
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations Mike Lewis on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
PowerHA/XD
- Thread starter spigven
- Start date
Hi
As a guide you need to
1. Define exactly what will run on each server , define your topology ( a schematic would be useful)
list cluster name , boot ip, virtual ip addresses , volume groups ( major numbers should be same on remote nodes)
2. Define size of LUNS to be used , make sure the same disks are accessible from the other nodes. Disk heartbeat LUN needs to be defined as well
3. Once your h/w topology is defined and all nodes can see each other, connect to each other , see the LUNS from all servers , then you can start creating your volume groups and filesystems
4. Then configure HACMP , i.e. topology first , then resource groups
5. Create your stop and start scripts
If the 3rd server would be for DR , I assume 2 nodes will have production and test environments and the 3rd node will only have production failed over onto it . if thats the case
if you have site B with its own SAN , you can mirror site A's LUNS to site B's at the OS level , plus you have to make sure whatever fails onto the site B node , is setup the same as site A's nodes.
Hope this helps
As a guide you need to
1. Define exactly what will run on each server , define your topology ( a schematic would be useful)
list cluster name , boot ip, virtual ip addresses , volume groups ( major numbers should be same on remote nodes)
2. Define size of LUNS to be used , make sure the same disks are accessible from the other nodes. Disk heartbeat LUN needs to be defined as well
3. Once your h/w topology is defined and all nodes can see each other, connect to each other , see the LUNS from all servers , then you can start creating your volume groups and filesystems
4. Then configure HACMP , i.e. topology first , then resource groups
5. Create your stop and start scripts
If the 3rd server would be for DR , I assume 2 nodes will have production and test environments and the 3rd node will only have production failed over onto it . if thats the case
if you have site B with its own SAN , you can mirror site A's LUNS to site B's at the OS level , plus you have to make sure whatever fails onto the site B node , is setup the same as site A's nodes.
Hope this helps
- Thread starter
- #3
Hi
Sorry for the late response , if you have 2 disk arrays one at each site , you can mirror LUNS at the Operating system level so that one disk/lun resides on array 1 at one site the other disk/lun on array 2 at the other site . Then use the Ooperating system to mirror the filesystems to these disks/luns. So if you lose a disk array , your apps will still be available
HTH
Sorry for the late response , if you have 2 disk arrays one at each site , you can mirror LUNS at the Operating system level so that one disk/lun resides on array 1 at one site the other disk/lun on array 2 at the other site . Then use the Ooperating system to mirror the filesystems to these disks/luns. So if you lose a disk array , your apps will still be available
HTH
- Status
Similar threads
- Locked
- Question