Are you duplicating the traffic over the same physical network by relaying through the VM, rather than writing directly to network storage, thereby halving the write performance? Assuming you're on a GigE network, are all the network devices running in full duplex? Just some guesses based on the fact that the throughput is almost exactly half. *Steve Dainard * IT Infrastructure Manager Miovision <http://miovision.com/> | *Rethink Traffic* 519-513-2407 ex.250 877-646-8476 (toll-free) *Blog <http://miovision.com/blog> | **LinkedIn<https://www.linkedin.com/company/miovision-technologies> | Twitter <https://twitter.com/miovision> | Facebook<https://www.facebook.com/miovision> * ------------------------------ Miovision Technologies Inc. | 148 Manitou Drive, Suite 101, Kitchener, ON, Canada | N2C 1L3 This e-mail may contain information that is privileged or confidential. If you are not the intended recipient, please delete the e-mail and any attachments and notify us immediately. On Thu, Sep 26, 2013 at 5:23 AM, Stefano Stagnaro <stefanos@prisma-eng.com>wrote:

Hello,

I'm testing oVirt 3.3 with GlusterFS libgfapi back-end. I'm using a node for engine and one for VDSM. From the VMs I'm mounting a second GlusterFS volume on a third storage server.

I'm experiencing very bad transfer rates (38MB/s) writing from a client to a VM on the mounted GlusterFS. On the other hand, from the VM itself I can move a big file from the root vda (libgfapi) to the mounted GlusterFS at 70MB/s.

I can't really figure out where the bottleneck could be. I'm using only the default ovirtmgmt network.

Thank you for your help, any hint will be appreciated.

Regards, -- Stefano Stagnaro IT Manager

Prisma Engineering S.r.l. Via Petrocchi, 4 20127 Milano – Italy

Tel. 02 26113507 int 339 e-mail: stefanos@prisma-eng.com skype: stefano.stagnaro

_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users