[01:24:39] *** Quits: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) (Ping timeout: 260 seconds) [01:54:44] *** Joins: tkulasek (~tkulasek@134.134.139.83) [02:05:26] *** Joins: tomzawadzki (tomzawadzk@nat/intel/x-zgimhfosceugonzf) [05:27:59] *** Quits: tkulasek (~tkulasek@134.134.139.83) (Ping timeout: 256 seconds) [05:30:55] *** Joins: tkulasek (~tkulasek@134.134.139.83) [05:46:50] *** Joins: haris_ (84ed9a7e@gateway/web/freenode/ip.132.237.154.126) [05:46:55] Hi [05:47:47] I was wondering if there is a way to test spdk (may be run the hello.c test) using some nvme emulator? and if there is a recommendation for an emulator.. [05:48:13] I have a VM running CentOS 7 on an Esxi.. [05:50:14] haris_: hey. I know you can emulate an NVMe drive in QEMU, I'm not sure about ESXi. [05:51:11] SPDK also contains a bdev layer that abstracts away all NVMe operations [05:51:56] so you could develop your application on top of a ramdisk, and then switch internally to NVMe when you're ready [05:52:31] darsto: Yes, I read about the emulation using qemu, so that would mean, I would use qemu to emulate nvme, and then run the SPDK in the VM (which is running through qemu)? [05:52:53] Yup, that's possible [05:54:07] Thank, and about the other way, using the bdev layer. It sounds like nvme spec emulation only using the ramdisk? Does that mean, I will be able to run all the Nvme commands and do simple IOs? [05:57:30] haris_: You can only issue raw NVMe commands to an NVMe device. SPDK bdev layer abstracts away raw device access and allows only generic I/O [06:00:51] Oh yes, I remember reading about bdev layer. So, won't that mean to use bdev layer, I would need access to a locally attached nvme device, or a remote nvme device? [06:03:24] there's a brief description of the SPDK bdev layer here - http://www.spdk.io/doc/bdev.html [06:05:26] the bdev layer offers a generic API for issuing I/O, so it doesn't make a difference if you're using NVMe/NVMe-oF/ramdisk/PMDK/etc under the hood [06:05:46] is that what you're asking about? I didn't quite catch that [06:10:25] Yes, so essentially, I can use the bdev layer to access blocks for IO, and internally I can configure SPDK to use anything (nvme, pmdk, logical vols). Sounds nice.. [06:15:28] Yes. If you're developing an application integrating SPDK, you might want to check out the Bdev programming guide - http://www.spdk.io/doc/bdev_pg.html [06:16:17] darsto: Thanks.. [06:46:08] *** Joins: erodriguez (d8f01e17@gateway/web/freenode/ip.216.240.30.23) [07:07:43] *** Parts: lhodev (~Adium@66-90-218-190.dyn.grandenetworks.net) () [07:07:48] *** Joins: lhodev (~Adium@66-90-218-190.dyn.grandenetworks.net) [07:46:22] FYI community meeting in 15 min. See http://www.spdk.io/community/ for WebEx info [08:34:32] *** Quits: tomzawadzki (tomzawadzk@nat/intel/x-zgimhfosceugonzf) (Ping timeout: 256 seconds) [09:07:59] Could somebody from Intel take a look at this and tell me ifs a spurious failure: https://ci.spdk.io/spdk/builds/review/e529500fa012aafad3e18494bd31c5e08b95c657.1521090041 [09:08:41] Tests in our lab are succeeding with this change so I'd like to understand if the failures are due to environmental differences. [09:14:17] hi Ed - it does look unrelated to your patch [09:15:03] I would re-run it, but it looks like your patch has a merge conflict and needs to be rebased anyway, so please look into that first, and then we can re-check it [09:30:15] *** Joins: travis-ci (~travis-ci@ec2-54-161-172-140.compute-1.amazonaws.com) [09:30:16] (spdk/master) blobstore: process split operation sequentially (Tomasz Zawadzki) [09:30:16] Diff URL: https://github.com/spdk/spdk/compare/a8d3ac0cf4b8...29e8013d0543 [09:30:16] *** Parts: travis-ci (~travis-ci@ec2-54-161-172-140.compute-1.amazonaws.com) () [09:45:52] *** Joins: boutcher_ (~boutcher@66.113.132.66) [09:47:39] *** Quits: boutcher (~boutcher@66.113.132.66) (Ping timeout: 260 seconds) [09:49:19] *** Joins: param (~param@106.206.67.111) [09:50:43] drv : jim : which is the next card that i can work on? Shall i start with that checking of the listen address in the list during connection.. [09:51:54] param: that would be great, yeah - it probably needs some test updates as well [09:52:43] okay..sure.. should a card be created in trello or not required? [09:53:57] it would be good to create a card just to track what is being worked on and avoid duplicated effort [09:57:37] sure.. [10:24:08] *** Quits: param (~param@106.206.67.111) (Remote host closed the connection) [10:38:09] *** Quits: tkulasek (~tkulasek@134.134.139.83) (Ping timeout: 248 seconds) [14:21:54] *** Joins: travis-ci (~travis-ci@ec2-54-91-233-154.compute-1.amazonaws.com) [14:21:55] (spdk/master) nvmf: retrieve namespace UUID from bdev (Daniel Verkamp) [14:21:55] Diff URL: https://github.com/spdk/spdk/compare/0bc4207ece3a...a05f88ff341d [14:21:55] *** Parts: travis-ci (~travis-ci@ec2-54-91-233-154.compute-1.amazonaws.com) () [14:26:25] *** Quits: mphardy (~mphardy@pool-72-83-7-2.washdc.fios.verizon.net) (Ping timeout: 246 seconds) [14:28:25] *** Joins: mphardy (~mphardy@pool-72-83-7-2.washdc.fios.verizon.net) [15:01:44] irc from 38000 feet [15:02:39] holy crap do i have a lot of patches to review! [15:07:12] :) [15:07:30] gerrithub at 3800 feet is not so great though [15:07:38] 38000 feet [15:08:22] how is the latency? [15:08:51] pretty bad - or at least not as good as my cube in chandler :) [16:13:59] *** Joins: travis-ci (~travis-ci@ec2-54-161-172-140.compute-1.amazonaws.com) [16:14:01] (spdk/master) bdev: Vbdev Passthru Example (paul luse) [16:14:01] Diff URL: https://github.com/spdk/spdk/compare/a05f88ff341d...21919851257f [16:14:01] *** Parts: travis-ci (~travis-ci@ec2-54-161-172-140.compute-1.amazonaws.com) ()