[01:02:24] *** Joins: tkulasek (tkulasek@nat/intel/x-ngskzkhvgpgzmuyu) [03:01:19] *** Joins: dlw1 (~Thunderbi@114.255.44.143) [03:03:20] *** Quits: dlw (~Thunderbi@114.255.44.143) (Ping timeout: 268 seconds) [03:03:21] *** dlw1 is now known as dlw [03:48:05] *** Quits: dlw (~Thunderbi@114.255.44.143) (Ping timeout: 240 seconds) [06:47:41] *** Joins: JoeGruher (c037362a@gateway/web/freenode/ip.192.55.54.42) [06:48:25] hi, i have SPDK NVMeoF target running and would like to try the SPDK NVMeoF initiator, I don't see it in the documentation, is it available and ready for testing? and where would I find setup instructions? [08:05:21] yes, it is working [08:07:30] see here http://www.spdk.io/doc/bdev.html#bdev_config_nvme [08:26:45] *** Joins: tomzawadzki (~tomzawadz@134.134.139.73) [10:10:38] *** Joins: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) [10:10:39] (spdk/master) env_dpdk: handle RTE_BAD_IOVA correctly in vtophys_get_paddr_pagemap (Jim Harris) [10:10:39] Diff URL: https://github.com/spdk/spdk/compare/08daa4477e8f...4d8c6a84407b [10:10:39] *** Parts: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) () [10:13:30] *** Joins: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) [10:13:31] (spdk/master) scripts: amend scheduler option in fio.py (Seth Howell) [10:13:31] Diff URL: https://github.com/spdk/spdk/compare/4d8c6a84407b...b484547f57c0 [10:13:31] *** Parts: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) () [10:15:58] *** Joins: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) [10:15:59] (spdk/master) conf: don't strdup name if section already exist (Pawel Wodkowski) [10:15:59] Diff URL: https://github.com/spdk/spdk/compare/c099f120cc7c...eac02a4ace1a [10:15:59] *** Parts: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) () [10:32:07] hmm, where did "destruct_split_vbdev" RPC come from? that isn't named consistently [10:32:36] looks like it is brand new, so I guess we can still change it [10:34:25] hmm, no, it was in the 18.04 release [10:40:47] *** Joins: travis-ci (~travis-ci@ec2-54-211-101-244.compute-1.amazonaws.com) [10:40:48] (spdk/master) bdev/split: free req in destruct_split_vbdev RPC call (Pawel Wodkowski) [10:40:49] Diff URL: https://github.com/spdk/spdk/compare/cc9c772379a2...4e012722efe9 [10:40:49] *** Parts: travis-ci (~travis-ci@ec2-54-211-101-244.compute-1.amazonaws.com) () [10:41:39] *** Quits: JoeGruher (c037362a@gateway/web/freenode/ip.192.55.54.42) (Ping timeout: 260 seconds) [11:59:49] *** Joins: JoeGruher (c0373628@gateway/web/freenode/ip.192.55.54.40) [11:59:59] can i define lvols in nvmf.conf.in? [12:02:20] also what might cause this: [12:02:23] rsa@tppjoe08:~/spdk/scripts$ sudo ./rpc.py construct_nvme_bdev -b spdkdev1 -t PCIe -a 0000:3d:00.0 Error while connecting to /var/tmp/spdk.sock Error details: [Errno 2] No such file or directory [12:28:07] hi JoeGruher [12:28:52] you cannot define lvols in nvmf.conf.in - lvols are based on on-disk metadata instead of configuration files or RPCs [12:29:05] do you have the nvmf_tgt application started? [12:29:45] *** Quits: tkulasek (tkulasek@nat/intel/x-ngskzkhvgpgzmuyu) (Ping timeout: 264 seconds) [12:49:16] jimharris: when you get a chance, please review this DPDK compatibility patch for RHEL/CentOS 6: https://review.gerrithub.io/#/c/spdk/dpdk/+/410559/ [13:07:30] *** Quits: JoeGruher (c0373628@gateway/web/freenode/ip.192.55.54.40) (Ping timeout: 260 seconds) [13:22:23] looks good to me - did you test it locally? [13:22:37] I tested it, but only on a newer system [13:22:48] once sethhowe is around, we should get the CentOS 6 machine back in the pool [13:23:18] he's only here for another day until he abandons us for a month :) [13:24:04] yeah [13:24:27] if he doesn't get to it, I'll dig around and see if I can figure out how the VM startup stuff works [14:02:30] *** Joins: dlw (~Thunderbi@114.252.37.68) [15:14:10] *** Quits: dlw (~Thunderbi@114.252.37.68) (Ping timeout: 264 seconds) [15:19:09] jimharris: I found a CentOS 6 VM that I have locally - let me test this before we check it in [15:31:09] *** Joins: MK_ (0fcbe94c@gateway/web/freenode/ip.15.203.233.76) [15:38:41] *** Joins: dlw (~Thunderbi@114.252.37.68) [16:04:05] *** Quits: dlw (~Thunderbi@114.252.37.68) (Ping timeout: 240 seconds) [16:04:20] *** Joins: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) [16:04:21] (spdk/master) include/string.h: add comments for public APIs (Yanbo Zhou) [16:04:22] Diff URL: https://github.com/spdk/spdk/compare/4e012722efe9...19354d966c9f [16:04:22] *** Parts: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) () [16:06:39] *** Joins: dlw (~Thunderbi@114.252.37.68) [16:09:00] *** Joins: travis-ci (~travis-ci@ec2-54-211-101-244.compute-1.amazonaws.com) [16:09:01] (spdk/master) rpc/python: Add abbreviated '-f' option to load_config (Shuhei Matsumoto) [16:09:02] Diff URL: https://github.com/spdk/spdk/compare/19354d966c9f...a3dd8df8cd5e [16:09:02] *** Parts: travis-ci (~travis-ci@ec2-54-211-101-244.compute-1.amazonaws.com) () [16:11:00] *** Joins: travis-ci (~travis-ci@ec2-54-159-95-249.compute-1.amazonaws.com) [16:11:01] (spdk/master) test/vhost: add vhost_nvme.c unit test framework (Chen Wang) [16:11:02] Diff URL: https://github.com/spdk/spdk/compare/a3dd8df8cd5e...d6d097a18d63 [16:11:02] *** Parts: travis-ci (~travis-ci@ec2-54-159-95-249.compute-1.amazonaws.com) () [16:14:39] *** Joins: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) [16:14:40] (spdk/master) nvme: apply Identify CNS quirk to NS ID Descriptor (Daniel Verkamp) [16:14:40] Diff URL: https://github.com/spdk/spdk/compare/d6d097a18d63...f750c4ebb553 [16:14:40] *** Parts: travis-ci (~travis-ci@ec2-54-81-162-236.compute-1.amazonaws.com) () [16:17:36] *** Quits: dlw (~Thunderbi@114.252.37.68) (Ping timeout: 268 seconds) [16:33:45] jimharris or bwalker: can you take a look at Shuhei's patches for blobstore sequence vs batch? https://review.gerrithub.io/#/c/spdk/spdk/+/411524/ [16:34:03] I seem to remember that part of the reason for using a sequence is to avoid generating too much I/O at once and exceeding the underlying device's queue depth [16:34:11] not sure if batch is smart enough to handle -ENOMEM responses and retry [16:38:07] *** Quits: tomzawadzki (~tomzawadz@134.134.139.73) (Ping timeout: 265 seconds) [16:43:42] it's not - i replied [16:44:41] unmap is the big problem - read/write I/O very rarely span a cluster boundary but it's not uncommon to unmap a huge blob [16:52:39] *** Quits: MK_ (0fcbe94c@gateway/web/freenode/ip.15.203.233.76) (Ping timeout: 260 seconds) [18:45:00] *** Joins: dlw (~Thunderbi@114.255.44.143) [22:26:59] *** Joins: dlw1 (~Thunderbi@114.255.44.143) [22:27:00] *** Quits: dlw (~Thunderbi@114.255.44.143) (Read error: Connection reset by peer) [22:27:00] *** dlw1 is now known as dlw