Re: [DRBD-user] One vs many DRBD resources

2011-09-26 Thread James R. Leu
On Mon, Sep 26, 2011 at 09:00:41AM +0200, Maciej Gałkiewicz wrote: > 2011/9/23 James R. Leu : > > We just went through this exercise and went with AOE instead > > of iSCSI because we found the current crop of iSCSI target and > > initiators to be lacking in terms selectively bringing a LUN 'up' > >

Re: [DRBD-user] One vs many DRBD resources

2011-09-26 Thread Maciej Gałkiewicz
2011/9/23 James R. Leu : > We just went through this exercise and went with AOE instead > of iSCSI because we found the current crop of iSCSI target and > initiators to be lacking in terms selectively bringing a LUN 'up' > and 'down'.  Specifically the iSCSI target implementations do not > release

Re: [DRBD-user] One vs many DRBD resources

2011-09-23 Thread James R. Leu
We just went through this exercise and went with AOE instead of iSCSI because we found the current crop of iSCSI target and initiators to be lacking in terms selectively bringing a LUN 'up' and 'down'. Specifically the iSCSI target implementations do not release the backstoring device when the LUN

[DRBD-user] One vs many DRBD resources

2011-09-23 Thread Maciej Gałkiewicz
Hello, I would like to implement DRBD with LVM volumes exported by iSCSI to virtual machines. Here (http://dl.dropbox.com/u/5820195/storage.svg) are two stacks which I am considering. I have tens of virtual machines. Each of them mounts 5GB partition(s) (the size may change) accessible through iSC