Riaan van Niekerk wrote:
It would be great if this could move past the ingredients phase ("use A, B and C") to the recipe phase ("first you ..."). There is definitely a huge demand for "pseudo-shared storage" clustering, based on what I have heard from my customers and what I have seen on this list.
I second that. The trouble that many people (like me) have when starting out is that it's difficult to justify dropping a relatively large amount of money on a storage array that might not work very well. Using GNBD it would surely be easier to prove the concept.
I remember a comment in a FAQ somewhere which said rather emphatically that GNBD could not do mirroring across two devices, and it suggested to go and look at Intermezzo or AFS. Two projects which appear to be very far indeed from being enterprise-ready.
-- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
