[00:53:56] *** Joins: tomzawadzki (uid327004@gateway/web/irccloud.com/x-xpmkqdeadesqbsfo) [01:21:52] *** Joins: travis-ci (~travis-ci@ec2-54-161-254-68.compute-1.amazonaws.com) [01:21:53] (spdk/master) nvmf/rdma: change the default buffer size. (Seth Howell) [01:21:54] Diff URL: https://github.com/spdk/spdk/compare/6134d778d46e...1d9be84bfdf8 [01:21:54] *** Parts: travis-ci (~travis-ci@ec2-54-161-254-68.compute-1.amazonaws.com) () [01:47:48] *** Joins: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) [08:04:47] *** Quits: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) (Read error: Connection reset by peer) [08:55:39] I'm not fully versed on the dynamic mem allocation provided in the later version(s) of the DPDK, so I have a question regarding that. If I specify '-s N' (where N > 0) at SPDK app startup with a DPDK version that supports dynamic memory, am I guaranteed that as long as the app makes it through early init, that it will have successfully been preallocated/reserved that amount of memory for its lifetime? [09:00:05] peluse: Regarding the 'io device already registered' prob, was that a result of a commit regression after 18.10 was cut? Or, did it exist prior to 18.10? Trying to grok what, if any, patch(es) related to the initial problem (and then the fix) will, or will not, be made part of 18.10.1. [09:05:15] lhodev: the "io device already registered" fixes will get integrated into 18.10.1 [09:05:43] jimharris: Thx Jim. Hey, sorry about missing your ping yesterday. [09:06:10] no worries - just wanted to bottom out on the Version and Source fields in the spec file [09:09:29] Cool. Saw that pwodkowx uploaded a new patch set. Got a couple responses about that. [09:10:55] Will add comments via Gerrit. [09:11:14] Meanwhile, can you address my other question (entirely unrelated) above about dynamic mem allocation? [09:36:41] lhodev: Yes - it will fail init if it cannot immediately obtain the amount of memory requested via the -s option [09:37:32] I still highly recommend that you pre-reserve memory if you are able to [09:45:52] Thx bwalker [10:12:05] *** Joins: travis-ci (~travis-ci@ec2-54-158-243-227.compute-1.amazonaws.com) [10:12:06] (spdk/master) test: update timing flags for vhost initiator (Seth Howell) [10:12:07] Diff URL: https://github.com/spdk/spdk/compare/1d9be84bfdf8...9c3b481b5ac7 [10:12:07] *** Parts: travis-ci (~travis-ci@ec2-54-158-243-227.compute-1.amazonaws.com) () [10:13:33] *** Joins: travis-ci (~travis-ci@ec2-54-83-106-104.compute-1.amazonaws.com) [10:13:34] (spdk/master) scripts/vagrant: add the ability to specify the nvme file. (Seth Howell) [10:13:34] Diff URL: https://github.com/spdk/spdk/compare/9c3b481b5ac7...b15ecfad85bd [10:13:34] *** Parts: travis-ci (~travis-ci@ec2-54-83-106-104.compute-1.amazonaws.com) () [10:49:48] *** Quits: tomzawadzki (uid327004@gateway/web/irccloud.com/x-xpmkqdeadesqbsfo) (Quit: Connection closed for inactivity) [10:53:18] *** Joins: travis-ci (~travis-ci@ec2-54-162-216-63.compute-1.amazonaws.com) [10:53:19] (spdk/master) thread: print dev->name in error messages (Jim Harris) [10:53:19] Diff URL: https://github.com/spdk/spdk/compare/b15ecfad85bd...f3d7215681dd [10:53:19] *** Parts: travis-ci (~travis-ci@ec2-54-162-216-63.compute-1.amazonaws.com) () [11:09:46] *** Joins: travis-ci (~travis-ci@ec2-54-161-254-68.compute-1.amazonaws.com) [11:09:47] (spdk/master) bdev/crypto: prevent duplicates from being added to global name list (Paul Luse) [11:09:48] Diff URL: https://github.com/spdk/spdk/compare/f3d7215681dd...094966811801 [11:09:48] *** Parts: travis-ci (~travis-ci@ec2-54-161-254-68.compute-1.amazonaws.com) () [11:21:12] *** Joins: travis-ci (~travis-ci@ec2-54-83-106-104.compute-1.amazonaws.com) [11:21:13] (spdk/master) thread: Update unit test to use new threading logic (Ben Walker) [11:21:13] Diff URL: https://github.com/spdk/spdk/compare/094966811801...3c9815082103 [11:21:13] *** Parts: travis-ci (~travis-ci@ec2-54-83-106-104.compute-1.amazonaws.com) () [11:21:58] *** Joins: travis-ci (~travis-ci@ec2-54-163-94-123.compute-1.amazonaws.com) [11:21:59] (spdk/master) nvme: add extended reservation data structure definition (Changpeng Liu) [11:21:59] Diff URL: https://github.com/spdk/spdk/compare/3c9815082103...d49bbbf05332 [11:21:59] *** Parts: travis-ci (~travis-ci@ec2-54-163-94-123.compute-1.amazonaws.com) () [11:45:18] nvme-of tcp is now public: https://review.gerrithub.io/#/c/spdk/spdk/+/425191/ [12:57:37] poking around github and noticed that github services are deprecated and will stop working in january [12:57:53] what are github services? [12:57:55] not a huge deal, but it does drive our IRC notifications [12:58:00] oh [12:58:00] Travis CI [12:58:07] wow [12:58:11] what are they going to use instead? [12:58:17] https://developer.github.com/changes/2018-04-25-github-services-deprecation/ [12:58:32] someone needs to move to one of these webhooks [12:59:13] oh, so Travis CI still works, it just works in a different way [12:59:22] i think so, if i'm reading this correctly [13:03:27] lhodev: could you try an experiment for me? [13:04:11] Sure [13:05:02] can you try the scheme described here to pick a different name for saving the tar file to the local system? [13:05:03] https://pagure.io/packaging-committee/issue/233 [13:05:19] also described here under the "Troublesome URLs" https://fedoraproject.org/wiki/Packaging:SourceURL [13:05:46] in this case it's not a "Troublesome URL" but we could use the same #/ scheme to pick an spdk- prefixed name for saving the tar file to the local system [13:06:24] Lemme have a read. Will get back with you later. [13:07:11] PSA: We had a couple of latent failures on the Jenkins pool this morning(related to the new VMs, not the previous QAT stuff). Those have been resolved, and I have retriggered a few builds, but please ping me if you had an issue with a test failing the RBD or lvol tests this morning, please ping me and I can reschedule it. [13:12:41] *** Joins: travis-ci (~travis-ci@ec2-54-158-243-227.compute-1.amazonaws.com) [13:12:42] (spdk/master) test: adjust scripts for 01.org qat driver (Seth Howell) [13:12:42] Diff URL: https://github.com/spdk/spdk/compare/d49bbbf05332...be5d98000f73 [13:12:42] *** Parts: travis-ci (~travis-ci@ec2-54-158-243-227.compute-1.amazonaws.com) () [13:40:20] jimharris: Unsure if I'm reading the docs correctly. I tried altering the spec file with the source line as follows: [13:40:27] Source: https://github.com/spdk/spdk/archive/v%{version}.tar.gz#/%{name}-%{version}.tar.gz [13:41:13] Is that how you would expect to specify it? [13:44:41] It doesn't appear to work. [13:45:09] Using spectool in an effort to download from the URL results in: [13:45:32] Getting https://github.com/spdk/spdk/archive/v18.10.1.tar.gz#/spdk-18.10.1.tar.gz to /home/lghartma/rpmbuild/SOURCES/spdk-18.10.1.tar.gz [13:45:42] curl: (22) The requested URL returned error: 404 Not Found [13:46:10] can you remove the .1? there is no 18.10.1 tag yet [13:46:38] Ah....let me retry. [13:48:12] Voila....with version as 18.10, that *did* work....retrieved v18.10.tar.gz and renamed it to spdk-18.10.tar.gz. [13:48:14] Fabulous [13:48:27] sweet [13:49:21] For completion sake, let me try it with the rpmbuild cmd [13:51:31] That worked too. Now, I have to approach our buildteam because they have special scripts that run to prepare for the build and sadly they are not visible to me so I cannot determine whether they'll work or not with this scheme. [14:03:08] You see, we maintain a fork/clone of the actual repo. So, when a build is initiated, there are special scripts that create the tarball from the repo using "git archive". How they parse and create the tarball name is something I'll have to research. [14:05:12] so you will need to make modifications to the spec file anyways? [14:36:20] *** Joins: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) [14:59:12] jimharris: Other than altering the Source line using the URL fragment as experimented with this afternoon, I don't *think* we'd need additional changes. [15:00:10] *** Joins: travis-ci (~travis-ci@ec2-184-73-35-63.compute-1.amazonaws.com) [15:00:11] (spdk/master) pkg: add spec file for RPM package build (Pawel Wodkowski) [15:00:11] Diff URL: https://github.com/spdk/spdk/compare/be5d98000f73...84c168fa6522 [15:00:11] *** Parts: travis-ci (~travis-ci@ec2-184-73-35-63.compute-1.amazonaws.com) () [15:00:49] I do need to talk to our internal buildteam and solicit feedback from them with respect to how they will parse that spec file to create the tarball, but I don't think that should impact what we do here upstream. [16:07:51] *** Joins: travis-ci (~travis-ci@ec2-54-145-203-220.compute-1.amazonaws.com) [16:07:52] (spdk/master) histograms: add unit tests for histogram merge (Piotr Pelplinski) [16:07:52] Diff URL: https://github.com/spdk/spdk/compare/84c168fa6522...8fb123553a69 [16:07:52] *** Parts: travis-ci (~travis-ci@ec2-54-145-203-220.compute-1.amazonaws.com) () [17:19:54] *** Quits: darsto (~darsto@89-78-174-111.dynamic.chello.pl) (Ping timeout: 240 seconds) [18:09:36] *** Joins: darsto (~darsto@89-78-174-111.dynamic.chello.pl)