[00:53:23] *** Quits: ziyeyang_ (~ziyeyang@192.55.54.41) (Quit: Leaving)
[01:18:25] *** Joins: tomzawadzki (~tomzawadz@192.55.54.41)
[01:46:22] *** Quits: baruch_ (~baruch@141.226.177.138) (Read error: Connection reset by peer)
[02:33:15] *** Quits: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) (Ping timeout: 260 seconds)
[02:51:42] *** Joins: baruch (~baruch@bzq-82-81-85-138.red.bezeqint.net)
[03:00:04] *** Parts: Param (~Param@106.51.65.217) ()
[03:11:25] *** Joins: spoddar (~spoddar@106.51.65.217)
[03:12:07] *** Joins: Param (~Param@106.51.65.217)
[03:12:18] hi
[03:12:42] hello
[03:47:54] *** Parts: Param (~Param@106.51.65.217) ()
[04:27:27] *** Quits: spoddar (~spoddar@106.51.65.217) (Quit: spoddar)
[04:50:28] *** Joins: Param (~Param@106.51.65.217)
[04:50:33] *** Parts: Param (~Param@106.51.65.217) ()
[05:35:36] *** Joins: spoddar (~spoddar@49.205.223.118)
[05:35:37] *** Quits: spoddar (~spoddar@49.205.223.118) (Remote host closed the connection)
[07:09:31] *** Quits: baruch (~baruch@bzq-82-81-85-138.red.bezeqint.net) (Ping timeout: 252 seconds)
[08:36:18] *** Joins: lhodev (~Adium@inet-hqmc05-o.oracle.com)
[08:37:52] *** Parts: lhodev (~Adium@inet-hqmc05-o.oracle.com) ()
[09:11:21] *** Quits: tomzawadzki (~tomzawadz@192.55.54.41) (Ping timeout: 240 seconds)
[09:37:35] when you guys have a minute, can you look at this series of patches from darsto? https://review.gerrithub.io/#/c/388505/
[09:37:39] it fixes the fio plugin problems
[09:37:48] and in general simplifies the virtio bdev
[10:08:33] I pushed the first two patches - this one needs to be rebased https://review.gerrithub.io/#/c/388835/
[10:08:40] but otherwise looks good
[10:09:26] I'll rebase for him - important to get that test turned back on
[10:10:41] and then this one would get abandoned? https://review.gerrithub.io/#/c/388985/
[10:13:11] that one, plus https://review.gerrithub.io/#/c/388698/ and https://review.gerrithub.io/#/c/388699/
[10:15:37] bwalker: i'm still here
[10:16:55] k, everything got rebased already
[10:17:01] yeah I just did it
[10:17:04] we'll merge shortly
[10:17:16] nice
[10:37:53] drv: could you take a look at these patches starting here? https://review.gerrithub.io/#/c/387690/
[10:38:11] would like to get these in since it enables some of the blobstore per-thread stuff we were talking about last week
[10:45:29] the 5th patch in that series failed, but I'm not sure why
[10:45:33] may be an intermittent
[10:45:39] the first 4 are good to go though
[10:45:48] blobstore doesn't need the 5th one anyway
[10:48:22] taking a look now
[10:49:41] *** Joins: smlsr (4132842a@gateway/web/freenode/ip.65.50.132.42)
[10:51:46] hello all - very very new to spdk (not to dpdk) - in any case looking to do a quick POC for some folks by enhancing spdk to use openssl and 256 AES encryption for ALL driver activity with a fixed key (so basically all reads and writes will be encrypted/decrypted with 1 static key) - where would be the easiest place to insert the required encryption/decryption call so all requests are handled? Again this is only a POC (looking to remove vo
[10:51:56] Any pointers would be greatly appreciated
[10:52:23] hi smlsr
[10:52:33] hello jim
[10:53:29] recommendation would be to create a new spdk bdev (block device) module that would intercept any read/write calls to another block device
[10:53:42] there are a few examples in the spdk tree that you can use for reference
[10:54:19] lib/bdev/split and lib/bdev/error are two examples
[10:54:51] bdev/split takes one block device and just splits it into a bunch of smaller block devices (not based on metadata or anything)
[10:55:04] bdev/error allows for injecting errors in the I/O patch to a specified block device
[10:56:29] interesting - let me look at "split" and see if I can just replicate that somewhat for MALLOC and a simple translated directory (eg: /test -> /encrypted/test) - that would be enough for the POC
[10:57:09] the purpose is merely to inject the ability encrypt/decrypt to a fs - and remove internal need for vormetric
[10:57:54] and support different back ends - thanks for the guidance
[12:15:19] bwalker: looks like your patches are failing due to something related to the virtio changes we pushed this morning
[12:15:19] http://spdk.intel.com/public/spdk/builds/review/2824efd35c5001b6f6dd50db18c5f8763d6af440.1511809022/
[12:15:30] http://spdk.intel.com/public/spdk/builds/review/6054acffe65396758947e3815c34b0c056279543.1511808432/
[12:22:37] *** Joins: baruch (~baruch@141.226.177.138)
[13:36:37] jimharris: any more patches that need to go in v17.10.1?
[13:37:38] it looks like Thomasz has a few more patches out for review on the v17.10.x branch
[13:43:03] jimharris: replied https://review.gerrithub.io/c/388765
[14:02:26] jimharris: I fixed it, it just has to run through the pool again and we can get it in
[14:09:10] *** Quits: smlsr (4132842a@gateway/web/freenode/ip.65.50.132.42) (Ping timeout: 260 seconds)
[15:50:15] *** Joins: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97)
[16:10:33] drv: first two patches here are ready for your re-review: https://review.gerrithub.io/#/c/387690/
[16:10:51] then we can get these first three patches in which should enable the blobstore threading work
[16:56:37] as long as I investigated about 4, 5 months ago.
[16:58:11] Hence we have to wait aio completion regardless of failure or success and we cannot agree with Ben's suggestion yet.
[16:58:36] Hi Jim, I reply to Gerrit too, thanks.
[17:00:13] hi Shuhei - so you also tried io_destroy()?
[17:01:23] if io_destroy() works, then we should not receive any completions for outstanding I/O, and can complete the bdev ios with abort status in the bdev aio driver
[17:02:29] unfortunately, io_destroy man page says "will block on the completion of all operations that could not be canceled"
[17:03:00] and if I recall correctly, io_cancel is not plumbed meaning this will be much worse than just waiting for all I/O to complete
[17:03:24] maybe
[17:07:37] bwalker: can you take another look at https://review.gerrithub.io/#/c/386837/ - Shuhei and I have both responded (it's the aio reset patch)
[20:01:17] *** Quits: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) (Quit: Page closed)