similar to: does libgfapi and fuse mount performance matches?

Dear support, I need to export gluster volume with LIO for a virtualization system. In this moment I have a very basic test configuration: 2x HP 380 G7(2 * Intel X5670 (Six core @ 2,93GHz), 72GB ram, hd RAID10 6xsas 10krpm, lan Intel X540 T2 10GB) directly interconnected. Gluster configuration is replica 2. OS is Fedora 27 For my tests I used dd and I found strange results. Apparently the

Good Evening, I have read that libgfapi has been backported to qemu-kvm in RHEL 6.5 (and by virtue CentOS and SL). However I am unable to figure out how to actually make it work as described. Virt-manager still only seems to support glusterfs volumes via fuse. I can use qemu-img to create a disk image on gluster://<server>/<Volume>. But virt-manager can only use it from a fuse

Hi, I created an encrypted volume which appears to be working fine with FUSE but the volume is supposed to store VM images (master key in place). I noticed some references in libgfapi source code to encryption so I decided to try it out. While attempting to create an image: # qemu-img create -f qcow2 gluster://gluster01/virt0/testing.img 30G Formatting

Hi, After updating glusterfs server to 3.8.5 (from Centos-gluster-3.8.repo) the KVM virtual machines (qemu-kvm-ev-2.3.0-31) that access storage using libgfapi are no longer able to start. The libvirt log file shows: [2016-11-02 14:26:41.864024] I [MSGID: 104045] [glfs-master.c:91:notify] 0-gfapi: New graph 73332d32-3937-3130-2d32-3031362d3131 (0) coming up [2016-11-02 14:26:41.864075] I [MSGID:

Hello. I'm playing with libgfapi network disks, over IB and all is working fine, but not disk locking (and true rdma transport). I use virtlockd, and with fuse mount, locking works as expected. But when i converted disk definitions to libgfapi, locks are not created (but qemu starts and works fine). I used direct and indirect locking - same result : qemu working fine, no locks. my

Hi, I am currently testing running Openstack instance on Cinder volume with libgfapi. This instance is the Windows instance and i found that when running random 4k write workload, the CPU utilization is very high, 90% CPU utilization with about 86% in privileged time. I also tested the workload with volume from NFS and the CPU utilization is only around 5%. For gluster fuse, the CPU utilization

On 08/11/2017 04:51 PM, Niels de Vos wrote: > On Fri, Aug 11, 2017 at 12:47:47AM -0400, Raghavendra Gowdappa wrote: >> Hi all, >> >> In a conversation between me, Milind and Csaba, Milind pointed out >> fadvise(2) [1] and its potential benefits to Glusterfs' caching >> translators like read-ahead etc. After discussing about it, we agreed >> that our

Dear All, Very pleased to see that the Redhat 6.5 beta promises "Native Support for GlusterFS in QEMU allows native access to GlusterFS volumes using the libgfapi library" Can I ask if virt-manager & libvirt can control libgfapi mounts? :) or do I need to use ovirt? :( many thanks Jake

Hey guys, I got another "reboot crash" with gfapi and this time libvirt-3.2.1 (from cbs.centos.org). Is there anyone who can audit the libgfapi usage in libvirt? :-) WK: I use bonded 2x10Gbps and I do get crashes only in heavy I/O situations (fio). Upgrading system (apt-get dist-upgrade) was ok, so this might be even related to amount of IOPS. -ps On Sun, Sep 10, 2017 at 6:37 AM, WK

On Fri, Aug 11, 2017 at 12:47:47AM -0400, Raghavendra Gowdappa wrote: > Hi all, > > In a conversation between me, Milind and Csaba, Milind pointed out > fadvise(2) [1] and its potential benefits to Glusterfs' caching > translators like read-ahead etc. After discussing about it, we agreed > that our performance translators can leverage the hints to provide > better

Hi, I've got a pair of systems running CentOS 7.4 as a testbed for kvm + gluster. It's a very basic config with a single SSD on each system. Gluster is configured on the two systems and I'm testing performance with fio. My test numbers directly against the brick and the the fuse mountpoint are really good. I started out creating a libvirt storage pool on the fuse mount and noticed

On Fri, Aug 11, 2017 at 05:50:47PM +0530, Ravishankar N wrote: > > > On 08/11/2017 04:51 PM, Niels de Vos wrote: > > On Fri, Aug 11, 2017 at 12:47:47AM -0400, Raghavendra Gowdappa wrote: > > > Hi all, > > > > > > In a conversation between me, Milind and Csaba, Milind pointed out > > > fadvise(2) [1] and its potential benefits to Glusterfs'

Hi list, We have an ancient hack that fuse not just passes on the statvfs data it's getting from the storage, but tweaks it by setting f_bsize / f_frsize to values of its own preference. [1] The supposed advantage is that f_bsize serves as a hint to applications for the preferred io size. (And regarding f_frsize -- in Linux it's a historical workaround for certain bugs in userspace[2]

I'm on 3.10.5. Its rock solid (at least with the fuse mount <Grin>) We are also typically on a somewhat slower GlusterFS LAN network (bonded 2x1G, jumbo frames) so that may be a factor. I'll try to setup a trusted pool to test libgfapi soon. I'm curious as to how much faster it is, but the fuse mount is fast enough, dirt simple to use, and just works on all VM ops such as

Hello, I'm having an issue with rebooting bricks holding images for live KVM machines (using libgfapi). I have a replicated+distributed setup of 4 bricks (2x2). The cluster contains images for a couple of kvm virtual machines. My problem is that when I reboot a brick containing a an image of a VM, the VM will start throwing disk errors and eventually die. The gluster volume is made like

hi, i am getting regular crashes which result in the mount being dropped: n1:~ # ls /n/auto/gv1/ ls: cannot access /n/auto/gv1/: Transport endpoint is not connected client side error log: http://pastebin.com/UgMaLq42 ..i am also finding that the gluster severs also sometimes just drop out - and i need to kill all the server side gluster processes and restart glusterd. i'm not sure if

So I noticed if I create files in the data brick path, the files travel to the other hosts too. Can I use the data brick path instead of using a fuse mount instead. I'm running two machines with two replicas. What happens if I do stripes? Some machines are clients as well as servers. Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL:

Hello all fellow GlusterFriends, I would like you to comment / correct my upgrade procedure steps on replica 2 volume of 3.7.x gluster. Than I would like to change replica 2 to replica 3 in order to correct quorum issue that Infrastructure currently has. Infrastructure setup: - all clients running on same nodes as servers (FUSE mounts) - under gluster there is ZFS pool running as raidz2 with SSD

Hi, I'm having an issue using the GlusterFS native client. After doing a mount the filesystem appears mounted but any operation results in a Transport endpoint is not connected message gluster peer status and volume info don't complain. I've copied the mount log below which mentions an error at fuse_init. The kernel is based on 2.6.15 and FUSE api version is 7.3. I'm using

+gluster-users Another guideline we can provide is to disable all performance xlators for workloads requiring strict metadata consistency (even for non gluster-block usecases like native fuse mount etc). Note that we might still can have few perf xlators turned on. But, that will require some experimentation. The safest and easiest would be to turn off following xlators: * performance.read-ahead