[02:08:45] *** Quits: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) (Ping timeout: 246 seconds) [02:46:00] *** Joins: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) [07:36:28] sethhowe: i'm sure you already know this, but looks like fedora-03 in the CTP is offline [07:44:20] *** Joins: travis-ci (~travis-ci@ec2-54-166-65-99.compute-1.amazonaws.com) [07:44:21] (spdk/v18.10.x) bdev/crypto: unregister io_device on failure in examine callback (Paul Luse) [07:44:21] Diff URL: https://github.com/spdk/spdk/compare/b177b5fcb829...126c22020a07 [07:44:21] *** Parts: travis-ci (~travis-ci@ec2-54-166-65-99.compute-1.amazonaws.com) () [07:45:16] *** Joins: travis-ci (~travis-ci@ec2-54-224-99-18.compute-1.amazonaws.com) [07:45:17] (spdk/master) check_format.sh: skip coding style checks if astyle is too old (Jim Harris) [07:45:17] Diff URL: https://github.com/spdk/spdk/compare/6c8af3ef7adf...760bb7e86249 [07:45:17] *** Parts: travis-ci (~travis-ci@ec2-54-224-99-18.compute-1.amazonaws.com) () [08:23:55] *** Joins: peluse (~peluse@134.134.139.72) [08:23:56] *** ChanServ sets mode: +o peluse [08:24:08] *** Joins: travis-ci (~travis-ci@ec2-54-205-44-160.compute-1.amazonaws.com) [08:24:09] (spdk/master) pmem: print logs on debug build (Pawel Wodkowski) [08:24:09] Diff URL: https://github.com/spdk/spdk/compare/760bb7e86249...2a8084018c93 [08:24:09] *** Parts: travis-ci (~travis-ci@ec2-54-205-44-160.compute-1.amazonaws.com) () [08:32:39] klateck, you there? [08:54:47] .join #jenkins [10:54:03] *** Joins: travis-ci (~travis-ci@ec2-54-166-107-244.compute-1.amazonaws.com) [10:54:04] (spdk/master) bdev/rbd: full control over client configuration (Patrick Ohly) [10:54:04] Diff URL: https://github.com/spdk/spdk/compare/38dfce0428ac...4cfae0360618 [10:54:04] *** Parts: travis-ci (~travis-ci@ec2-54-166-107-244.compute-1.amazonaws.com) () [10:55:01] *** Joins: travis-ci (~travis-ci@ec2-54-205-44-160.compute-1.amazonaws.com) [10:55:02] (spdk/v18.10.x) bdev: unregister bdevs top-down during shutdown. (Andrey Kuzmin) [10:55:02] Diff URL: https://github.com/spdk/spdk/compare/126c22020a07...77f89cbc3161 [10:55:02] *** Parts: travis-ci (~travis-ci@ec2-54-205-44-160.compute-1.amazonaws.com) () [11:21:52] jimharris, for compress UT, I assume you're good with having the compress vbdev UT and reduce UT share the same test flag meaning reduce will only get tested on TP machines where compression is tested? [11:23:12] sounds good to me [11:23:41] cool [11:24:06] working on getting at least the framework for compress vbdev UT up shortly [14:17:30] Hi there peluse [14:17:34] I dropped in for a sec [14:27:50] klateck, wow, pretty late there! [14:28:13] Yeah, wanted to check for a reply from Vishal [14:31:03] so what's up? [14:31:36] Is it about Jenkins mail thread I just saw?:) [14:32:30] Looks very promising [14:33:16] see prv irc [15:53:35] peluse: this configshell ubuntu bug - did you only see it on ubuntu 14.04? [15:53:39] or 16 as well? [15:54:27] its on both the ubuntu machines in the CH TP [15:54:56] doing 3 thigns at once or I would have just gone and fixed it, didn't want to forget so opened the issue... [15:55:26] no worries - wasn't sure if you're using ubuntu, so i was going to push a fix [15:55:42] do you have a link you can point me to? there are 3 ubuntu systems in the CTP [15:58:46] i found it [15:58:50] 16.04 failure looks different though [15:58:57] out of space error? [16:00:18] sethhowe: what happened to the tmux session on spdk-ci-003? [16:00:37] i was going to try to check on ubuntu-16.04 myself [16:04:30] I put a link to the failure in the GH issue [16:05:06] I didn't actually look at the 2nd ubunutu failure though [16:07:03] Nothing, I am messing around in it. [16:07:29] jimharris: sorry about that. What were you going to check on? I was looking at something for Paul [16:09:35] Jimharris: Just catching up on the logs. Yeah ubuntu16 was completely out of space on the /boot partition. I cleaned it up though. It looks like the only failure is on ubuntu14.04 in the pip3 installations. [16:19:00] *** Quits: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) (Quit: My Mac Pro has gone to sleep. ZZZzzz…) [16:19:23] ok - i'm going to change it to stop using pip and install the python packages directly [16:19:32] and then also skip trying to install configshell/pexpect on ubuntu 14 [16:19:38] similar to checks on rhel [16:37:30] thanks guys - certainly didn't mean to cause ints here for anyone just didn't want to forget about it later.... [16:48:38] jimharris: sounds good to me. I got into quite the rabbit hole trying to figure out why pip3 wasn't working on ubuntu 14.04 [17:26:42] *** Joins: travis-ci (~travis-ci@ec2-54-198-227-242.compute-1.amazonaws.com) [17:26:43] (spdk/master) reduce: don't pass size when loading pmem file (Jim Harris) [17:26:43] Diff URL: https://github.com/spdk/spdk/compare/4cfae0360618...1d75aad3ff9d [17:26:43] *** Parts: travis-ci (~travis-ci@ec2-54-198-227-242.compute-1.amazonaws.com) () [17:34:33] *** Joins: travis-ci (~travis-ci@ec2-54-204-168-176.compute-1.amazonaws.com) [17:34:34] (spdk/master) Boxes to support libvirt as provider. Virtualbox set as default. (Lavar Askew) [17:34:34] Diff URL: https://github.com/spdk/spdk/compare/1d75aad3ff9d...1cbd14ca1bb4 [17:34:34] *** Parts: travis-ci (~travis-ci@ec2-54-204-168-176.compute-1.amazonaws.com) () [18:14:46] oh man, what took 20 min at work to attempt to download/build jenkins took just barely 60 sec at home.... [18:15:29] wont be able to finish testing changes to the gerrit plugin probably until Sun though.... maybe Mon [18:49:56] *** Quits: klateck (~klateck@134.134.139.72) (Ping timeout: 272 seconds) [18:49:57] *** Quits: jimharris (~jimharris@134.134.139.72) (Ping timeout: 272 seconds) [20:39:06] *** Joins: jimharris (~jimharris@134.134.139.72) [20:39:06] *** ChanServ sets mode: +o jimharris [20:45:48] *** Joins: travis-ci (~travis-ci@ec2-54-205-44-160.compute-1.amazonaws.com) [20:45:48] (spdk/master) env: make the generic device attach API public (Darek Stojaczyk) [20:45:49] Diff URL: https://github.com/spdk/spdk/compare/1cbd14ca1bb4...38462248bf64 [20:45:49] *** Parts: travis-ci (~travis-ci@ec2-54-205-44-160.compute-1.amazonaws.com) () [21:08:58] *** Joins: LiuXiaodong (c037362d@gateway/web/freenode/ip.192.55.54.45) [21:23:58] jimharris: Looking over the changes in gerrithub marked as part of branch v18.10.x, did all of the patches regarding the nvmf_tgt setfaulting get backported (yet). I recall sethhowe: mentioning that they were hashtag'd as 18.10.1 at one point. [21:24:43] not sure - bwalker is back in the office tomorrow - will sync up with him and sethhowe [21:27:56] *** Joins: shadiasd (18cf3aad@gateway/web/freenode/ip.24.207.58.173) [21:37:39] Hi there! I am looking for a "blktrace" equivalent tool that can be used on SSDs with SPDK setup. I need to trace the IO pattern on my SSD, but blktrace seems to not be an option when using SPDK. [21:38:34] I was wondering if anyone has ever done IO access pattern tracing on a SSD with SPDK setup? [21:39:00] thank you in advance for any help! [21:40:04] On the GerritHub GUI (either Old UI or New UI), I still don't see how I am unable to remove the Chandler-Test-Pool vote. Is my only option to re-run through the test pool via asking a maintainer or making an alteration (commit msg, rebase, etc.) and re-push ? [21:40:40] s/unable/able/ [21:50:43] lhodev: the only way right now is one of the options you listed - you can't remove it directly [21:51:52] it's a bit of a pain to allow everyone to remove it, i'm not aware of a way in gerrit to say that anyone can remove the CTP's +1, but not someone else's +1 [21:52:20] we plan to retire the Chandler Test Pool soon, so then this problem will go away [21:52:31] Ok. I thought I had heard during a meeting in the past that we could click on something in the UI. Guess I misunderstood (or was dreaming). I just rebased and pushed. [21:52:32] shadiasd: good question [21:53:08] are you using the spdk bdev layer? or just the nvme driver? [21:53:39] Keeping fingers crossed Seth identifies something squirrely about fedora-03 before my job runs ;-) [21:53:54] sethhowe+ [21:54:42] shadiasd: there is some event tracing functionality built into the nvme-of target, the iscsi target and the bdev layer currently - it's still a bit of a WIP but you can get a snapshot of bdev io elapsed times at a nanosecond level with what's there today [21:55:49] it's not plumbed into the nvme driver (yet) - you can get some more details on enabling it at https://spdk.io/doc/nvmf.html [22:23:46] @jimharris, thank you for the reply! I am just using the nvme driver, very unlucky :) In this case, is there any way to get a simple trace that shows 1) IO addresses that are being accessed 2) IO command (read or write)? Assuming that we do not care a bout the exact timing and only care about the the ordering. [22:25:06] I see that SPDK has a trace library (include/spdk/trace.h). Can this library be potentially used to get the IO trace, when using the nvme driver?