From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============6822118863876104950==" MIME-Version: 1.0 From: Sreeni (Sreenivasa) Busam (Stellus) Subject: Re: [SPDK] Need help for fixing NVMe probe problem in NVMeoF initiator running fio. Date: Thu, 22 Mar 2018 17:52:18 +0000 Message-ID: In-Reply-To: 1513875044.2658.19.camel@intel.com List-ID: To: spdk@lists.01.org --===============6822118863876104950== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Ben, Thank you for your reply. I fully forgot about the email thread, but I got = the fio working and able to get my project related to developing a tool for= IOPS and latency completed. The problem was device was removed by the admi= nistrator and I got it working. My apologies for not getting back. I have few questions about how to handle I/O errors in NVMF initiator, am g= oing to start a different thread. Sreeni -----Original Message----- From: SPDK [mailto:spdk-bounces(a)lists.01.org] On Behalf Of Walker, Benjam= in Sent: Thursday, December 21, 2017 8:51 AM To: spdk(a)lists.01.org Subject: Re: [SPDK] Need help for fixing NVMe probe problem in NVMeoF initi= ator running fio. On Wed, 2017-12-20 at 21:31 +0000, Sreeni (Sreenivasa) Busam (Stellus) wrot= e: > Initiator log: > ./fio --name=3Dnvme --numjobs=3D1 --filename=3D"trtype=3DRDMA adrfam= =3DIPV4 > traddr=3D172.17.2.175 trsvcid=3D11345 subnqn=3Dnqn.2017-06.io.spdk-MPcnod= e1 = > ns=3D1" -- bs=3D4K --iodepth=3D1 -- = > ioengine=3D/home.local/sfast/spdk20/spdk/examples/nvme/fio_plugin/fio_pl > ugin -- > sync=3D0 --norandommap --group_reporting --size=3D12K --runtime=3D3 = > -rwmixwrite=3D30 > --thread=3D1 --rw=3Drw > nvme: (g=3D0): rw=3Drw, bs=3D(R) 4096B-4096B, (W) 4096B-4096B, (T) = > 4096B-4096B, ioengine=3Dspdk, iodepth=3D1 > fio-3.3 > Starting 1 thread > Starting DPDK 17.11.0 initialization... > [ DPDK EAL parameters: fio -c 0x1 -m 512 --file-prefix=3Dspdk_pid2821= 4 ] > EAL: Detected 32 lcore(s) > EAL: No free hugepages reported in hugepages-1048576kB > EAL: Probing VFIO support... > nvme_ctrlr.c:1031:nvme_ctrlr_construct_namespaces: *ERROR*: = > controller has > 0 namespaces > fio_plugin.c: 298:spdk_fio_setup: *ERROR*: spdk_nvme_probe() This error means your NVMe-oF subsystem does not have any namespaces. That = usually means there is a problem in the configuration file. > NVMF configuration file: > [Nvmf] > MaxQueuesPerSession 4 > AcceptorPollRate 10000 > = > [Subsystem1] > NQN nqn.2017-06.io.spdk-MPcnode1 > Core 1 > SN SPDK0000000000000001 > Listen RDMA 172.17.2.175:11345 > AllowAnyHost Yes > NVMe 0000:84:00.0 What version of SPDK are you using? The NVMe directive inside of the [Subsy= stem] section was replaced with syntax that looks like this: Namespace Nvme0n1 1 And Nvme0n1 is defined by the [Nvme] section with syntax like this: [Nvme] TransportId "trtype:PCIe traddr:0000:00:00.0" Nvme0 See etc/spdk/nvmf.conf.in for a full example. If you upgraded SPDK to a new= er version but didn't update your configuration file, you'll end up with a = subsystem with 0 namespaces because it doesn't understand the NVMe directiv= e anymore. --===============6822118863876104950==--