Re: Controller health check | |
We use RDAC driver, which creates the dar, dac, hdisk, utm devices etc.
Ok, Since other devices e.g fcs1 etc also use RDAC, So no need to remove driver, let me put it in other way ..How do i need to reconfigure these drivers again for new HBA? So that they keep the same multipath, policies of failover etc..
In our setup we use fastt 700, shared storage with HACMP, active-passive cluster. Failed HBA is on active node. There are two servers in cluster, with two DAR named dar0, dar1, two DAC dac0, dac2, shared VGs, LVs are created across both to provide resilience under HACMP.
I read something relating Fastt700, will the following will be sufficient
1. HACMP resource groups, move them to failing over node, ensure VGs are varied off
2. Since all the RGs are moved to the other node, now all the devices on the redundant on the faulty HBA and can remove fcs0, fscsi0, dar0, dar1 , hdisk*, using rmdev
3. Run the cfgmgr -v
After that do i need to export the VGs or the ODM entries for VG, disks etc remains same...
5. move the RGs back
Rgds
Yash  |