[00:30:06] *** Joins: mszwed (~mszwed@134.134.139.72) [01:31:34] *** Joins: tkulasek (~tkulasek@134.134.139.72) [03:47:59] *** Joins: ksiero (c1de87c2@gateway/web/freenode/ip.193.222.135.194) [03:49:08] Hello, we were thinking about setting thin allocated volumes as two legs of raid1 setup [03:49:41] however, when the raid synchronises, the zeroes would inflate the other volume [03:50:29] what happens when writes that include only zeroes is written to non-allocated logical volume sectors ? [03:50:56] does the space get allocated or no? [04:29:41] *** Quits: ksiero (c1de87c2@gateway/web/freenode/ip.193.222.135.194) (Ping timeout: 256 seconds) [08:37:31] jimharris, lhodev: how urgent is spdk-examples package? Can we go without out-of-tree build? [08:42:22] do you mean ignore an spdk-examples package for now - just focus on lib and app? [08:43:02] yes [08:43:03] examples has nvme/perf and nvme/identify which are both really useful - but the rest ranges from nice-to-have to not required at all [08:43:47] we can install examples and source for them but they will not build. [08:43:52] i'm fine with just focusing on lib and app for right now - do you see any major blockers that would preclude spdk-examples in the future? [08:44:07] I think the hello* examples would be up there on the list as well [08:44:34] the hello* examples are only useful if you have source though - by themselves they don't do anything [08:45:00] oh duh, yes of course [08:45:39] paluse: nope, without full source any application fail to build becouse it try to build everything and of cource we don't want to include full spdk source in spdk-examples package. [08:45:40] maybe we need some new top level directory (utils?) for things like nvme/perf and nvme/identify [08:46:36] jimharris: we can install binaries and source but rebuilding will fail [08:47:41] so, spdk_example_xxx binary will be roking when spdk-examples package is installed and source bill be available for it but thats it. [08:48:29] are we fine with that or we can only include binaries redirecting to github for source [08:49:22] * roking -> working :D [08:52:50] * peluse likes roking better :) [08:54:34] what is the downside of having the source available in a package? [08:59:58] hmmm.. I have 2 dev systems both running Fed27 and I use samba on both to connect to my repos from my mac. Same exact smb.conf file on each and one works like a champ, the other consistently sets +x permissions when I save a file. Not a huge deal especially now since check)format reminds me to fix it but curious if anyone has any ideas on what to look at? [09:01:39] jimharris: only downside is possible question: "Why you examples are not building?" [09:04:30] forgive my ignorance - why would the examples not build from source? [09:07:48] paluse: https://unix.stackexchange.com/a/103418 [09:09:32] jimharris: currently examples will try to build libraries as they depend on them, but examples package will contain only source of examples. [09:11:05] so the examples source package would include the top-level Makefile? [09:11:07] we should have some switch in out build system to tell the makefile "we need only to buid app and link with system installed libraries including -lspdk*" [09:11:44] only mk/* and examples/* but [09:13:06] it doesn't matter as lib folder wont be there [09:13:18] make[2]: *** No rule to make target '/usr/share/spdk/build/lib/libspdk_event_bdev.a', needed by 'hello_bdev'. St [09:15:28] peluse: https://unix.stackexchange.com/a/103418 [09:25:20] pwodkowx, saves me again, thanks! For this machine I needed this for some reason: [Global] [09:25:20] [09:25:20] map archive = no [09:25:20] [09:25:29] (was in the webpage you pointed me to) [09:45:36] *** Quits: tkulasek (~tkulasek@134.134.139.72) (Quit: Leaving) [10:24:01] *** Joins: Tracy35 (~Tracy35@12.218.82.130) [10:26:16] Would like to know if nvmef target can handle nvme smart-log command sent from the host? [10:40:38] You mean log identifier 0x2 for the Get Log Page command? [10:42:32] this is the code right now: [10:42:34] case SPDK_NVME_LOG_HEALTH_INFORMATION: [10:42:34] /* TODO: actually fill out log page data */ [10:42:34] return SPDK_NVMF_REQUEST_EXEC_STATUS_COMPLETE; [10:42:52] in lib/nvmf/ctrlr.c around line 1100 [10:43:21] so it won't crash, but it doesn't actually attempt to get the information from the backing storage device and populate the log page currently [10:43:37] (patches accepted) [11:21:27] Thank you, @bwalker [11:33:13] pwodkowx: jimharris: Sorry to be late to the party on packaging examples. I was away for a bit. My opinion: I'd really like to see at least some of the examples like "perf" and "identify" executables installed as one of the spdk-XXX packages for the value they bring independent of learning the code through example. With a small tweak to the build -- just adding an install target -- for those, I think we can get there pretty [11:33:13] easily. [11:45:34] *** Quits: lhodev_ (~lhodev@66-90-218-190.dyn.grandenetworks.net) (Quit: Textual IRC Client: www.textualapp.com) [11:46:07] *** Joins: lhodev (~lhodev@66-90-218-190.dyn.grandenetworks.net) [11:52:56] *** Quits: Tracy35 (~Tracy35@12.218.82.130) (Ping timeout: 268 seconds) [11:55:03] *** Joins: travis-ci (~travis-ci@ec2-54-166-177-65.compute-1.amazonaws.com) [11:55:04] (spdk/master) configure: Move checks for nasm and ipsec (Seth Howell) [11:55:04] Diff URL: https://github.com/spdk/spdk/compare/5c263046dbe6...2b6d9ce21f7c [11:55:04] *** Parts: travis-ci (~travis-ci@ec2-54-166-177-65.compute-1.amazonaws.com) () [12:03:25] *** Joins: travis-ci (~travis-ci@ec2-54-81-49-114.compute-1.amazonaws.com) [12:03:26] (spdk/master) memory: register each hotplugged page separately (Darek Stojaczyk) [12:03:26] Diff URL: https://github.com/spdk/spdk/compare/2b6d9ce21f7c...f4cb6c90b4d8 [12:03:26] *** Parts: travis-ci (~travis-ci@ec2-54-81-49-114.compute-1.amazonaws.com) () [13:09:09] *** Joins: travis-ci (~travis-ci@ec2-54-166-120-10.compute-1.amazonaws.com) [13:09:10] (spdk/master) env_dpdk: Eliminate unneeded start group ld opt (Lance Hartmann) [13:09:10] Diff URL: https://github.com/spdk/spdk/compare/f4cb6c90b4d8...8337cc2d0cd0 [13:09:10] *** Parts: travis-ci (~travis-ci@ec2-54-166-120-10.compute-1.amazonaws.com) () [13:12:26] *** Joins: travis-ci (~travis-ci@ec2-54-156-92-210.compute-1.amazonaws.com) [13:12:27] (spdk/master) jsonrpc-client: add new C client library (Liu Xiaodong) [13:12:27] Diff URL: https://github.com/spdk/spdk/compare/8337cc2d0cd0...c10f8e160e42 [13:12:27] *** Parts: travis-ci (~travis-ci@ec2-54-156-92-210.compute-1.amazonaws.com) () [13:49:56] *** Joins: travis-ci (~travis-ci@ec2-54-197-134-208.compute-1.amazonaws.com) [13:49:57] (spdk/master) pkgdep.sh: Fix error handling for ipsec submodule (Seth Howell) [13:49:57] Diff URL: https://github.com/spdk/spdk/compare/e3cd058cb46f...5faf653f0c4b [13:49:57] *** Parts: travis-ci (~travis-ci@ec2-54-197-134-208.compute-1.amazonaws.com) () [13:50:43] *** Joins: travis-ci (~travis-ci@ec2-54-159-14-68.compute-1.amazonaws.com) [13:50:44] (spdk/master) nvme: Don't alias sgl1 when building request (Ben Walker) [13:50:44] Diff URL: https://github.com/spdk/spdk/compare/c10f8e160e42...e3cd058cb46f [13:50:44] *** Parts: travis-ci (~travis-ci@ec2-54-159-14-68.compute-1.amazonaws.com) () [13:52:58] *** Joins: travis-ci (~travis-ci@ec2-54-159-14-68.compute-1.amazonaws.com) [13:52:59] (spdk/master) scripts/perf: fix nvme perf scripts (Karol Latecki) [13:52:59] Diff URL: https://github.com/spdk/spdk/compare/5faf653f0c4b...1250a41664d8 [13:52:59] *** Parts: travis-ci (~travis-ci@ec2-54-159-14-68.compute-1.amazonaws.com) () [13:54:02] *** Joins: travis-ci (~travis-ci@ec2-54-81-49-114.compute-1.amazonaws.com) [13:54:03] (spdk/master) rte_vhost: include for CentOS-6 (wuzhouhui) [13:54:03] Diff URL: https://github.com/spdk/spdk/compare/ddb17216a7fc...5e2a99e508d9 [13:54:03] *** Parts: travis-ci (~travis-ci@ec2-54-81-49-114.compute-1.amazonaws.com) () [13:59:00] *** Joins: travis-ci (~travis-ci@ec2-54-196-85-132.compute-1.amazonaws.com) [13:59:01] (spdk/master) lib/event: update the check of poller returned status (GangCao) [13:59:01] Diff URL: https://github.com/spdk/spdk/compare/5e2a99e508d9...3daa9e6f756f [13:59:01] *** Parts: travis-ci (~travis-ci@ec2-54-196-85-132.compute-1.amazonaws.com) ()