Distributed network RAID

[Bart Silverstrim]

Sean Carolan wrote:
>> How robust is this setup in your experience?  Quirks?  Problems?
>>
>> And how difficult is it to maintain or recover from a problem with this?
> 
> I set up linux-HA and drbd on an NFS share in a VMWare lab, and yanked the 
> virtual power on one box, then watched it fail over, then went back and forth 
> several times in this manner.  Reading or writing to the mounted share took a 
> little longer while it was doing it's failover, but no data was lost. 
> Development seems robust, and they have quite a few success stories:

Any issues with the replication, i.e., machine A dies, use machine B for 
the day, bring machine A back online...did you have to do anything to 
get the data to sync, or was it more or less automatic, or was there 
something like in the old days of NT domains where once the backup was 
"promoted" to primary there really wasn't a way to demote it?

Also most of what I'd found said that if you have a shared resource 
machine A would have the drive as read/write, while machine b was 
read-only.  Is that the way it is with this?