On 04/06/2017 03:22 AM, yipik...@gmail.com wrote: > On 06/04/2017 09:42, Nick Fisk wrote: >> >> I assume Brady is referring to the death spiral LIO gets into with >> some initiators, including vmware, if an IO takes longer than about >> 10s. I haven’t heard of anything, and can’t see any changes, so I >> would assume this issue still remains. >> >> >> >> I would look at either SCST or NFS for now. >> > LIO-TCMU+librbd-iscsi [1] [2] looks really promising and seams to be the > way to go. It would be great if somebody as insight about the maturity > of the project, is it ready for testing purposes ? >
It is not mature yet. You can do IO to a rbd image, but it currently does a queue depth of only 1. We are in the process of merging patches from a couple branches to add rbd aio support, failover/failback across gateways, perf improvements, and lots of bug fixes. With them, linux works well, and we are working on a couple windows bugs. For ESX, we are hoping to be ready around the end of summer. You should not use ESX with tcmu/tcmu-runner right now, because several commands are not implemented or implemented incorrectly for ESX. _______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
