Ever been in a situation where you have plenty of 16x DVD-R media but your laptop’s slimline DVD writer has a maximum write speed of 8x?
I’m sure you’ve been there. For most users having a DVD to write slower than usual is not a big deal if they’re just creating a DVD for an odd occasion or two. (I mean, with USB’s flash drives and hard drives with an storage capacity that’s about 210x of what a DVD can store, who even bothers with optical drives anymore?)
If you’re just the average user, just stop here, move along nothing exciting to see. 😉
If you’re the user who wants the best burning speed possible, well read on. Or just have a really good optical drive that does an excellent job in ripping audio CD’s, read on too.
Apparently there’s a storage protocol called iSCSI. You may needed it at some point to provide a form of shared storage carved out of your LVM pool or ZFS pool for your high availability clusters. Anyhow, by using the Linux-iSCSI stack (configured with the targetcli shell), you can literally share an optical drive to the network at the raw device level. (Hint: use the pscsi backstore inside targetcli, you can’t export an optical drive via the iblock backstore, it doesn’t work that way)
What does raw device level mean to the average user? It means that a Linux machine with a recent kernel (3.2+) can share it’s optical drives with any machines that have an iSCSI initiator. Couple that with a Gigabit Ethernet connection, you can finally write DVD media at the full speed of what your media is capable of.
Oh by the way, you can also have your virtual machines running any virtualization platform to create optical media reliably regardless if the hypervisor supports writing to optical media from a virtual machine.
Anyways just a few words of caution, if you do decide to use multiple DVD burners on a single iSCSI target, you could do that but not with EAC. EAC requires a different IQN per optical drive. If you have multiple LUNs on a single IQN, EAC will only recognize the optical drive on LUN 0. I’m not sure if this is a bug with EAC or an issue with the iSCSI initiator.