[00:51:00] *** Joins: tkulasek (~tkulasek@134.134.139.72) [01:34:51] *** Joins: webchat75 (65e6cf8e@gateway/web/freenode/ip.101.230.207.142) [01:35:14] hello [01:36:02] is any one here? [01:42:04] webchat75: Please go ahead and ask your question. If someone is around that knows the answer, they will reply. [01:42:40] thanks [01:43:22] is there any small task can i do about SPDK [01:45:26] webchat75: I'm not directly involved in SPDK development, so I'm not the best person to answer this, but check out the Trello boards: https://trello.com/spdk [01:45:51] webchat75: It includes a "Things To Do" board with "Low Hanging Fruit" (smaller/easy tasks) [01:46:07] That might be a good starting point if you want to get involved. [01:47:38] ok,thanks. [03:04:40] *** Quits: webchat75 (65e6cf8e@gateway/web/freenode/ip.101.230.207.142) (Ping timeout: 260 seconds) [03:05:54] *** Quits: mszwed (mszwed@nat/intel/x-ixwswxazgbkmxiwm) (Ping timeout: 272 seconds) [03:06:48] *** Joins: mszwed (mszwed@nat/intel/x-jtwxczxtzqwuobeh) [03:39:58] *** Quits: dlw (~Thunderbi@114.255.44.143) (Ping timeout: 264 seconds) [03:59:39] *** Quits: kjakimia (kjakimia@nat/intel/x-avoqpwehgracxozi) (Ping timeout: 260 seconds) [03:59:45] *** Joins: kjakimia_ (kjakimia@nat/intel/x-lswtgqyqauhmmzvn) [04:00:09] *** kjakimia_ is now known as kjakimia [05:42:27] drv, so ony my strange QAT kernel driver problem, Tushar fixed it with 'depmod -a' [05:43:02] stefanha, good answer! :) [05:55:56] *** Joins: dlw (~Thunderbi@114.246.88.80) [06:00:11] *** Quits: dlw (~Thunderbi@114.246.88.80) (Ping timeout: 246 seconds) [06:49:35] *** Joins: dlw (~Thunderbi@114.246.88.80) [07:25:28] *** Quits: dlw (~Thunderbi@114.246.88.80) (Ping timeout: 268 seconds) [07:32:26] *** Joins: dlw (~Thunderbi@114.246.88.80) [08:04:27] *** Quits: dlw (~Thunderbi@114.246.88.80) (Ping timeout: 240 seconds) [08:05:55] *** Joins: tomzawadzki (~tomzawadz@134.134.139.72) [08:06:45] *** Quits: tomzawadzki (~tomzawadz@134.134.139.72) (Client Quit) [08:06:48] *** Joins: tzawadzki (~tomzawadz@134.134.139.72) [08:20:38] *** Joins: lhodev (~lhodev@inet-hqmc06-o.oracle.com) [08:54:21] *** Quits: tkulasek (~tkulasek@134.134.139.72) (Ping timeout: 255 seconds) [09:55:56] *** Joins: travis-ci (~travis-ci@ec2-54-92-251-103.compute-1.amazonaws.com) [09:55:57] (spdk/master) test/vhost: fix the test suite name (Chen Wang) [09:55:57] Diff URL: https://github.com/spdk/spdk/compare/6cd524d87c3a...885160c08b83 [09:55:57] *** Parts: travis-ci (~travis-ci@ec2-54-92-251-103.compute-1.amazonaws.com) () [09:58:26] *** Joins: travis-ci (~travis-ci@ec2-54-80-126-136.compute-1.amazonaws.com) [09:58:27] (spdk/master) vhost/nvme: release the dev before exit (Chen Wang) [09:58:27] Diff URL: https://github.com/spdk/spdk/compare/885160c08b83...6cfddbc71d21 [09:58:27] *** Parts: travis-ci (~travis-ci@ec2-54-80-126-136.compute-1.amazonaws.com) () [10:02:06] *** Joins: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) [10:02:07] (spdk/master) bdev/iscsi: silence clang enum conversion warning (Daniel Verkamp) [10:02:07] Diff URL: https://github.com/spdk/spdk/compare/6cfddbc71d21...87295c3c2c4c [10:02:07] *** Parts: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) () [10:02:15] *** Joins: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) [10:02:16] (spdk/master) nvmf: support transport data block SGL in get_xfer (Daniel Verkamp) [10:02:16] Diff URL: https://github.com/spdk/spdk/compare/60232f67e41f...432e5f655e90 [10:02:16] *** Parts: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) () [10:07:14] darsto: looking at the dynamic memory patch now [10:07:19] this is going to be so great for SPDK [10:07:51] it looks like Anatoly incorporated all of my feedback - the rte_mem_event_callback_register and rte_memseg_contig_walk functions were added just for us [10:16:08] jimharris: any thoughts on darsto's patch that disables the iSCSI hotplug tests? https://review.gerrithub.io/#/c/spdk/spdk/+/410176/ [10:16:20] where are we on fixing iSCSI hotplug? was someone actively working on that? [10:21:40] bwalker: did you see Shuhei's comment on your first perf patch? https://review.gerrithub.io/#/c/spdk/spdk/+/391876/ [10:21:49] *** Joins: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) [10:21:50] (spdk/master) iscsi: Add JSON config dump for iSCSI subsystem (Shuhei Matsumoto) [10:21:50] Diff URL: https://github.com/spdk/spdk/compare/432e5f655e90...abfed2298133 [10:21:50] *** Parts: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) () [10:23:06] *** Joins: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) [10:23:07] (spdk/master) bdev/nvme: check passthru cmd size against MDTS (Daniel Verkamp) [10:23:07] Diff URL: https://github.com/spdk/spdk/compare/abfed2298133...2aced8c74c6a [10:23:07] *** Parts: travis-ci (~travis-ci@ec2-54-211-38-28.compute-1.amazonaws.com) () [10:24:22] yeah I need to add a "per namespace" to the help statement [10:24:49] I mean, we don't actually clarify for the queue depth that it's per queue either [10:25:19] but I'll clean it up [11:15:35] drv: i added +2 to darsto's patch [11:21:42] *** Joins: travis-ci (~travis-ci@ec2-23-20-242-232.compute-1.amazonaws.com) [11:21:43] (spdk/master) ut/bdev_mt: add test case for ENOMEM handling with multiple io targets (Dariusz Stojaczyk) [11:21:43] Diff URL: https://github.com/spdk/spdk/compare/2aced8c74c6a...bfb73837de0d [11:21:43] *** Parts: travis-ci (~travis-ci@ec2-23-20-242-232.compute-1.amazonaws.com) () [12:48:20] *** Joins: tomzawadzki (~tomzawadz@134.134.139.72) [12:48:20] *** Quits: tzawadzki (~tomzawadz@134.134.139.72) (Remote host closed the connection) [13:36:25] jimharris: Is there ever a case during examine or something where spdk_bdev_open will get called with write permission on a device that is already claimed? Which then of course fails [13:36:54] I'm seeing this occur during the lvol tests all over the place - I think the blobstore is trying to open Nvme0n1 to read metadata, but it fails because something else has already claimed it [13:37:15] and when it fails, everything in the test works just fine - it handles the failure [13:37:31] but with Andrey's patch that lets you call spdk_bdev_open regardless of claim, it blows up [13:47:22] do you know what else has claimed it? the split module perhaps? [13:48:02] I don't now, but about to find out [14:02:45] *** Quits: lhodev (~lhodev@inet-hqmc06-o.oracle.com) (Remote host closed the connection) [14:15:01] yeah, so split claims it [14:15:06] then blobstore tries to read it [14:15:28] but blobstore tries to examine it by opening it with a write descriptor [14:15:38] I think blobstore examine needs to be read-only, then promote [14:19:18] *** Joins: travis-ci (~travis-ci@ec2-23-23-0-116.compute-1.amazonaws.com) [14:19:19] (spdk/v18.04.x) blob: don't try to claim cluster 0 in recovery code (Jim Harris) [14:19:19] Diff URL: https://github.com/spdk/spdk/compare/f1b747f50f38...5f59d919a4f0 [14:19:19] *** Parts: travis-ci (~travis-ci@ec2-23-23-0-116.compute-1.amazonaws.com) () [14:27:27] the read-only examine doesn't work because we mark the super block dirty immediately [14:27:31] so that write fails [14:30:19] what do you think the best solution for dealing with the split claim vs. metadata claim issue"? [14:36:28] we need to break the examine process into two parts [14:36:45] all gets back to that thread we had on the github issue yesterday [14:37:11] yeah - do you think we expose two different examine callbacks? [14:37:14] or add a flag? [14:37:14] blobstore needs to be fixed to not do a write during load [14:37:45] two examine callbacks might be better to cover the case where a module might need both [14:38:05] for example, a striping module that can either stripe based on DDF metadata or based on some explicit config file specification [14:39:36] blobstore also needs to save the size (or at least the number of clusters) of the underlying disk in its superblock - i don't think it does that today [14:40:51] *** Quits: tomzawadzki (~tomzawadz@134.134.139.72) (Ping timeout: 255 seconds) [14:41:31] so who's the unlucky person getting an email from me explaining what they need to do to fix this? [14:41:49] * jimharris rolls dice [14:42:01] the blobstore fixes can be separate from the examine fixes [14:42:20] my magic 8-ball says ppelplin gets the honor [14:42:26] for which one? blobstore? [14:42:45] i think there's three - roughly in this order: [14:42:50] 1) examine callbacks [14:42:56] 2) don't write during load [14:43:05] 3) save cluster/size in super block [14:43:29] ok - and you want him on all 3? I think the first one is independent [14:43:33] do you agree with that order? [14:43:40] no - I was thinking just #1 [14:43:46] I think it's #1 off to the side, then #2 and #3 separately [14:43:58] 3 depends on 2, but 2 doesn't depend on 1 [14:44:03] true [14:44:18] let me sync up with you offline [14:44:22] ok [15:01:20] *** Joins: tomzawadzki (tomzawadzk@nat/intel/x-giopamzdbxjcgobh) [16:34:23] *** Joins: travis-ci (~travis-ci@ec2-54-91-238-210.compute-1.amazonaws.com) [16:34:24] (spdk/master) nvmf: Keep track of outstanding requests (Ben Walker) [16:34:24] Diff URL: https://github.com/spdk/spdk/compare/bfb73837de0d...7740b02e36ee [16:34:24] *** Parts: travis-ci (~travis-ci@ec2-54-91-238-210.compute-1.amazonaws.com) () [16:41:40] FYI Asia time friendly Community meeting coming up in a little over 4 hours. See http://www.spdk.io/community/ for WebEx info and https://trello.com/b/DvM7XayJ/spdk-community-meeting-agenda for agenda topics [17:02:22] *** Quits: tomzawadzki (tomzawadzk@nat/intel/x-giopamzdbxjcgobh) (Ping timeout: 256 seconds) [18:23:44] *** Joins: dlw (~Thunderbi@114.255.44.143) [18:32:00] *** Joins: dlw1 (~Thunderbi@114.255.44.143) [18:32:01] *** Quits: dlw (~Thunderbi@114.255.44.143) (Read error: Connection reset by peer) [18:32:01] *** dlw1 is now known as dlw [18:41:55] *** Joins: webchat75 (65e6cf8e@gateway/web/freenode/ip.101.230.207.142) [21:06:16] *** Joins: bwalker_ (~bwalker@ip70-190-226-244.ph.ph.cox.net) [21:06:16] *** ChanServ sets mode: +o bwalker_ [23:00:41] *** Joins: dlw1 (~Thunderbi@114.255.44.143) [23:00:41] *** Quits: dlw (~Thunderbi@114.255.44.143) (Read error: Connection reset by peer) [23:00:42] *** dlw1 is now known as dlw [23:47:49] *** Quits: bwalker_ (~bwalker@ip70-190-226-244.ph.ph.cox.net) (Quit: Leaving)