[09:16:24] *** Joins: bwalker (~bwalker@134.134.139.83) [09:16:24] *** ChanServ sets mode: +o bwalker [09:16:25] *** Server sets mode: +cnrt [09:16:25] *** Server sets mode: +cnrt [09:21:32] *** Joins: sethhowe (~sethhowe@134.134.139.83) [09:31:37] *** Joins: travis-ci (~travis-ci@ec2-35-168-21-7.compute-1.amazonaws.com) [09:31:37] (spdk/master) blob: Don't look at cluster map prior to checking frozen (Jim Harris) [09:31:37] Diff URL: https://github.com/spdk/spdk/compare/5316b12e5e98...e740ba637c02 [09:31:37] *** Parts: travis-ci (~travis-ci@ec2-35-168-21-7.compute-1.amazonaws.com) () [10:36:33] *** Joins: travis-ci (~travis-ci@ec2-54-80-195-78.compute-1.amazonaws.com) [10:36:33] (spdk/master) nvmf: Improve nvmf_lvol.sh (Ben Walker) [10:36:33] Diff URL: https://github.com/spdk/spdk/compare/e740ba637c02...219b62430db0 [10:36:33] *** Parts: travis-ci (~travis-ci@ec2-54-80-195-78.compute-1.amazonaws.com) () [10:36:58] *** Joins: travis-ci (~travis-ci@ec2-3-80-193-53.compute-1.amazonaws.com) [10:36:58] (spdk/master) test/qos: disable iscsi qos tests (Darek Stojaczyk) [10:36:58] Diff URL: https://github.com/spdk/spdk/compare/219b62430db0...e600f0967665 [10:36:58] *** Parts: travis-ci (~travis-ci@ec2-3-80-193-53.compute-1.amazonaws.com) () [12:00:08] sethhowe, klateck: are there limits on the size of the build artifacts? i'm having problems with my rocksdb patch that gzips a few tracepoint files into the output directory [12:00:10] https://10.102.17.104:8080/job/RBD_BlobFS_tests/job/BlobFS_autotest/24284/console [12:00:24] this same problem has hit twice in a row - i'm sure it's related to my patch [12:00:32] https://review.gerrithub.io/c/spdk/spdk/+/452263 [12:14:19] jimharris wow, first time I see smth like that. To be honest I'm not aware of such limitations [12:14:36] About what size are we talking about here? GBs? [12:19:53] I found this just now: https://issues.jenkins-ci.org/browse/JENKINS-47566 [12:20:35] And looking at the files in the workspace on the node which last failed, there are in fact files which are owned by root and their permissions are 600. [12:21:12] Could you try adding chmod at the end of your script? [13:29:11] klateck - thanks! yeah, this looks like the problem, thanks for figuring this out [14:32:33] lib/nvmf/transport.c:88 - is this check still valid? [14:32:58] i've ended up in a rabbit hole and am hitting this issue [14:33:17] 1) create transport w/ -u 8192 [14:33:32] then nvmf tgt adjusts the io unit size based on the devices max i/o size - so makes it 45056 [14:33:42] we save the config, reload, and then this error hits on line 88 [14:41:16] *** Quits: gila (~gila@94-212-217-121.cable.dynamic.v4.ziggo.nl) (Quit: My Mac Pro has gone to sleep. ZZZzzz…) [15:05:47] bwalker: any update on the forthcoming availability of a paper comparing Jens Axboe's io_uring with SPDK? [15:56:01] *** Joins: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) [16:18:59] All failed due to rocksdb version. [16:19:27] lhodev: Not yet, but work is still active [16:20:02] we had some results at the summit that I think showed io_uring around 1.5M random reads per second per core [16:20:02] The current strategy is to update nightly test from 19.01 to 19.04 and rocksdb fix will not be merged into 19.01 ? [16:20:13] and SPDK at 8+M [16:20:36] It's OK to me but failures in more than 10 patches is noisy. [16:20:53] jimharris is sorting out rocksdb at the moment I think [16:21:34] hi Shuhei - are you referring to the nightly test failures? [16:21:51] No, the patches for iscsi. [16:22:24] iSCSI had some issues, and I fixed them but all were for the master. [16:22:36] But jrlusby has used 19.01. [16:22:57] So I rebased the patches onto 19.01 manually. But they all failed due to rocksdb. [16:23:15] They are just noisy. [16:23:17] ah - yeah there is a patch coming for the 19.01.x branch to fix that [16:23:25] Yes. [16:23:25] ok - let me try merging this patch to 19.01.x [16:23:27] we needed to version rocksdb appropriately [16:23:42] Thank you! [16:31:00] I think 3 patches starting here should do it: [16:31:02] https://review.gerrithub.io/c/spdk/spdk/+/452482 [16:40:06] @jimharris: thank you I review them and try to rebase my patches onto them. [17:28:41] @jimharris: https://review.gerrithub.io/#/c/spdk/spdk/+/452482/ looks the patch failed in the same test case as before. Log was not collected and so I retriggerd. [17:30:31] bwalker: Thx for the update. [17:31:38] Not critical but fixing it may be helpful for future rocksdb updates. I just shared. [17:34:05] Shuhei were planning on upgrading to 19 04 as you recommended [17:34:16] Don't worry too much about 19 01 for our sake [17:34:38] We are* [17:45:01] So your upgrading is very welcome. [18:43:29] *** Quits: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) (Ping timeout: 256 seconds)