similar to: [Gluster-devel] Poor performance of block-store with RDMA

On Wed, Sep 13, 2017 at 1:03 PM, GiangCoi Mr <ltrgiang86 at gmail.com> wrote: > Hi all > Hi GiangCoi, The Good news is that now we have gluster-block [1] which will help you configure block storage using gluster very easy. gluster-block will take care of all the targetcli and tcmu-runner configuration for you, all you need as a pre-requisite is a gluster volume. And the sad part is

[ Hi Takao. Could you attach some logs which we can diagnostic? On 2017? 08? 15? 19:42, Hatazaki, Takao wrote: > > Hi, > > > > I did ib_write_lat in perftest. It worked fine. Between servers and > between server and client, 2-byte latency was ~0.8us, 8MB bandwidth > was ~6GB/s. Very normal with IB/FDR. > > > >

Hello Ji-Hyeon, Thanks, is that option available in 3.12 gluster release? because we're still on 3.8 and just playing around latest version in order to have our solution migrated. Thank you! 9/21/17 2:26 PM, Ji-Hyeon Gim ?????: > Hello Pavel! > > In my opinion, you need to check features.shard-block-size option first. > If a file nobigger than this value, it would not be

Hi all I want to configure glusterfs to expose iSCSI target. I followed this artical https://pkalever.wordpress.com/2016/06/23/gluster-solution-for-non-shared-persistent-storage-in-docker-container/ but when I install tcmu-runner. It doesn't work. I setup on CentOS7 and installed tcmu-runner by rpm. When I run targetcli, it not show *user:glfs* and *user:gcow* */>* ls o- /

Hi Ji-Hyeon, Thanks to your help I could find out the problematic file. This would be the quota file of my volume it has a different checksum on node1 whereas node2 and arbiternode have the same checksum. This is expected as I had issues which my quota file and had to fix it manually with a script (more details on this mailing list in a previous post) and I only did that on node1. So what I now

On 2017? 08? 06? 15:59, mabi wrote: > Hi, > > I have a 3 nodes replica (including arbiter) volume with GlusterFS > 3.8.11 and this night one of my nodes (node1) had an out of memory for > some unknown reason and as such the Linux OOM killer has killed the > glusterd and glusterfs process. I restarted the glusterd process but > now that node is in "Peer Rejected"

Hi, I'm getting system (and eventually cluster) crashes on intensive disk writes in ubuntu server 10.04 with my OCFS2 file system. I have an iSER (infiniband) backed shared disk array with OCFS2 on it. There are 6 nodes in the cluster, and the heartbeat interface is over a regular 1GigE connection. Originally, the problem presented itself while I was doing performance testing and

Out of haste I decided for the first time not to pull down and compile iet to export a single volume to a windows machine so I just used tgtd. Well, much more time later than otherwise would be the case had I used iet, when I mount the volume on the windows ini, it can't find and install a driver? WTF? This have anything to do with the tgtd being a tech preview:) Anybody seen this b4? jlc

+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

Hi Xen-Users! I ask for the infiniband support of the xen hypervisor as describet in the following papers http://www.openfabrics.org/archives/spring2007sonoma/Monday%20April%2030/Xiong%20OFA-Sonoma-2007-04-30-SoftIB.ppt and http://www.*xen*.org/files/*Xen*_RDMA_Voltaire_YHaviv.pdf The papers read that a direct infiniband support is possible to circumvent the dom0 to access infiniband devices

This patchset includes an updated version of the SCSI frontend and backend drivers. The frontend and backend drivers exchange SCSI RDMA Protocol (SRP) messages via a ring buffer. The backend driver sends SCSI commands to the user-space daemon, which performs SCSI commands and I/O operations. The backend driver uses VM_FOREIGN feature like the blktap driver for zero-copy of data pages. Like the

Hello all. We recently purchased HP Proliant DL380 G7 server and also a HP Lefthand P4500 G2 24TB iSCSI box. The server itself has Quad Ethernet based on the BCM5709 Chipset from Broadcom (NetXtreme II). My main question is about iSCSI offload mechanism, as iSCSI itself is pretty new to me but I have gotten a good hold on it tho. I have managed to setup and configured our Lefthand box into

Hello everyone, We're currently working with using libvirt as an abstract API to make our dealings with other hypervisors far easier and faster, however, we have our own storage API that connects/disconnects and makes LUNs available over iSCSI with iSER. We would have loved to use a storage pool however our system implements a one target per LUN/VDI, therefore, we'd have to create a pool

hi everyone, how can I use a LV created by lxc in libvirt guest domain? You know, you lxc-create with LV as backstore for a container(ubuntu in my case) and now you want to give to libvit. That, I thought would be a very common case but I failed to find a good howto on how one: lxc-create(lvm) => let libvirt take control over it. Or it's not how you would do it at all? many thanks, L.

.. what is? fellow centosians. how to you mount your loopback targets? I'm trying lvm backstore, I was hoping I would do it with uuid, but it's exposed more than once and how would kernel decide which device to use I don't know. thanks

Hi, I was wondering if anyone has successfully configured two lsi/3ware 9750-4i series controllers for multipathing under CentOS 5.7 x86_64? I've tried some basic setups with both multibus and failover settings, and had repeatable filesystem corruption over a iscsi(tgtd) or nfs3 connection. Any ideas? Vahan

Hey everyone, Is anybody aware of a /true/ active/active multi-head and multi-target clustered iSCSI daemon? IE: Server 1: Hostname: host1.test.com IP Address: 10.0.0.1 Server 2: Hostname: host2.test.com IP Address: 10.0.0.2 Then they would utilize a CLVM disk between them, let's call that VG "disk" and then directly map each LUN (1,2,3,4,etc) to LV's named 1,2,3,4,... and

On Mon, Apr 11, 2016 at 9:14 PM, Nathan Coulson <nathan at bravenet.com> wrote: > Hello > > We were attempting to use scsi-target-utils, hosted on a xen dom0 vm using > localhost, and running into some problems. I was not able to reproduce this > on a centos 7.2 server using the default kernel. Have you tried booting the Virt SIG kernel natively and seeing if you can

On 2016-04-12 09:43 AM, Nathan Coulson wrote: > By natively, I take it using > kernel /vmlinuz (vs kernel /xen) > > Not yet, but working on setting up such an environment. > > (At this time, I was using virt-install to reproduce the problem, and > the original server we are testing on did not support kvm but the 2nd > server does). > > On 2016-04-12 03:26 AM, George

This idea is intruiging... Suppose one has a set of file servers called A, B, C, D, and so forth, all running CentOS 6.5 64-bit, all being interconnected with 10GbE. These file servers can be divided into identical pairs, so A is the same configuration (diks, processors, etc) as B, C the same as D, and so forth (because this is what I have; there are ten servers in all). Each file server has