12:02:54 #startmeeting Weekly community meeting 27/Jul/2016 12:02:54 Meeting started Wed Jul 27 12:02:54 2016 UTC. The chair is kshlm. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:02:54 Useful Commands: #action #agreed #halp #info #idea #link #topic. 12:02:54 The meeting name has been set to 'weekly_community_meeting_27/jul/2016' 12:03:13 He was probably napping. 12:03:19 #topic Roll call 12:03:29 * pranithk1 here 12:03:30 * kkeithley is here 12:03:34 * ankitraj here 12:03:40 * hagarth is here 12:03:44 I'll be waiting for a minute to the agenda open. 12:03:50 * samikshan is here 12:03:58 * karthik_ is here 12:04:07 * jiffin is here 12:04:20 hagarth, pranithk1, You guys are here after a long long long time! 12:04:21 * rjoseph is here 12:04:29 * ramky is here 12:05:14 * ndevos arrives, after a short sprint 12:05:14 Okay. Good attendence today. 12:05:20 Welcome everyone! 12:05:30 #topic Next weeks meeting host 12:05:36 * anoopcs is also here 12:05:41 First up we select the host for the next weeks meeting. 12:05:43 kshlm: taking some time off from my school dropoff schedule ;) 12:05:48 Any volunteers? 12:06:02 hagarth, :) 12:06:09 can a intern be volunteer 12:06:13 ? 12:06:40 Sure. 12:06:55 * ankitraj interested 12:07:10 Just watch how we do it. 12:07:16 And do the same next week. 12:07:29 ankitraj: we're all part of the Gluster community, we dont care about job titles here ;-) 12:07:37 kshlm: ok 12:07:54 ankitraj, So you're the next host? 12:08:15 yea 12:08:18 Okay. 12:08:31 #info ankitraj is next weeks meeting host. 12:08:55 #topic GlusterFS 4.0 12:09:10 Let's have some updates now. 12:09:51 atinm, jdarcy, are you guys here? 12:09:57 Yes. 12:10:05 Okay. Good. :) 12:10:22 Not sure if it's still considered 4.0, but multiplexing is coming along nicely. 12:10:44 I'm at the point where basic stuff works and I'm running real regression tests to see which fail. 12:10:55 nice! 12:11:21 Patch still at http://review.gluster.org/#/c/14763/ 12:11:47 That's it, pretty much. Nothing on JBR, and I doubt there will be for a while. 12:11:53 Cool! 12:12:46 Work on GD2 is going on slowly. 12:13:01 ppai got a few fixes merged this week. 12:13:16 I'm still working on the transaction RPCs. 12:13:48 Anyone have updates on DHT2? 12:14:05 If not let's move onto the next topic. 12:14:49 #topic GlusterFS-3.9 12:15:05 pranithk1, Do you have any updates to share with us? 12:15:18 kshlm: yes 12:15:31 kshlm: We have 3 more features we need to add to 3.9 schedule 12:15:47 kshlm: reset-brick, block and object storage related ones 12:16:11 kshlm: Oh forgot one more, i.e. EC encoding perf improvement patch 12:16:28 kshlm: We will be updating the doc by end of this week... 12:16:36 kshlm: I mean the roadmap 12:16:40 Thanks pranithk1. 12:16:44 oh, and md-cache improvements from poornima 12:17:08 ndevos: Oh cool, more the merrier 12:17:25 maybe Kerberos support too, but that is mostly in the hands of Csaba, and I do not know how much time he can spend on it 12:17:31 pranithk1, Are we going to do anything about sub-directory mounts? 12:17:42 Jose and I will update roadmap shortly w/ common-HA improvements to include Samba 12:17:43 kshlm: use NFS! :P 12:17:53 kkeithley: cool. 12:18:04 kshlm: I think we should talk about it next week... 12:18:29 pranithk1, We should have talked about 3 weeks back. 12:18:39 But I've been out of office. 12:18:41 kshlm: :-). Well better late than never I guess 12:18:54 Let's see if we can get together earlier. 12:18:59 kshlm: yeah 12:19:03 Alright. 12:19:14 kshlm: that's all from me 12:19:21 I'll move on to the next topic then. 12:19:33 #topic GlusterFS-3.8 12:19:42 ndevos, The stage is yours. 12:19:50 Also jiffins 12:20:08 nothing exciting to mention, and 3.8 should be very boring at the moment anyway 12:20:44 I'm asking in many backports if they are really needed, we want bugfixes and stability improvements only 12:21:50 Ah, I knew I'd forgotten something. I need to send the PR for the update release-process and guidelines. 12:21:55 for example, the token-bucket-filter patch to move the functions from bitrot to libglusterfs should probably not be backported, instead *real* backports for eventual bugfixes would be a better approach 12:22:08 yes, I was going to ask about it :) 12:22:28 #action kshlm to send PR for release process document update 12:23:05 and if my firefox would respond, I'd add a link to the patch where the tbf discussion is taking place... 12:24:16 #link http://review.gluster.org/14945 12:25:07 Sorry. I'd hoped my new internet connection wouldn't be giving trouble. 12:25:15 But apprantly it does. 12:25:42 #info do not backport unneeded changes, save your time for working on other things and do not risk instability of the 3.8 release :) 12:26:02 So ndevos, 3.8.2 is on track for the 10th then? 12:26:24 yes, there is no reason to change that atm 12:26:34 Good to know. 12:26:53 I'll move on to the next topic if we have nothing more to discuss about 3.8 12:27:04 nothing from me 12:27:07 #topic GlusterFS-3.7 12:27:25 So I guess I'm the 3.7.14 release manager. 12:27:55 congrats! 12:28:11 Yay me! 12:28:54 As of now, 3.7.14 is on target for the 30th. 12:29:11 I'll be making an announcement of this later today. 12:30:38 the 30th what ? 12:31:17 30th of July. 12:31:27 this month, thats the thing with monthly updates :) 12:31:48 There have been 18 changes merged on top of 3.7.13. 12:32:23 A good amount IMO. 12:32:34 And that's it about 3.7. 12:32:53 still some patches pending? 12:33:21 * ndevos tries to trick firefox in showing the search results, but it takes a while... 12:33:23 I see a few more. 12:33:23 kshlm: I have one patch pending in EC... 12:33:32 kshlm: I will be sending it by EOD 12:33:34 pranithk1, You have until the 30th. 12:33:39 To get it merged. 12:33:57 Also, remember to add the bug to the tracker bug. 12:33:58 kshlm: yes sir 12:34:05 Cool! 12:34:54 ndevos, There have been ~15 reviews for 3.7 that have had activity since 3.7.13 was released. 12:35:11 Onto 3.6 12:35:19 #topic GlusterFS 3.6 12:35:47 I sent an update to the -users and -devel lists on the status of 3.6. 12:36:00 thanks kshlm! 12:36:46 So we're now officially not going to be doing any more stability releases for 3.6. 12:37:13 3.6 will get releases for security or very critical issues. 12:37:29 But only till 3.9 is released. 12:37:50 #link https://www.gluster.org/pipermail/gluster-devel/2016-July/050256.html 12:38:01 And that's about it for 3.6. 12:38:16 Onto the related projects. 12:38:22 #topic NFS Ganesha 12:39:08 we broke building nfs-ganesha against the glusterfs/master branch with http://review.gluster.org/#/c/14702/ :-( 12:39:59 I'll be doing some libgfapi corrections for that, and jiffin wil modify Ganesha when I'm done 12:40:23 How was the breakage detected? 12:40:31 Was it an automated test? 12:40:43 by trying to setup night builds for nfs-ganesha... 12:40:49 *nightly even 12:40:49 Cool! 12:41:31 hurray for CI 12:41:35 but, it should have been obvious if the patch was reviewed by the gfapi maintainers, they normally pay attention to these minor things 12:42:30 ndevos, Did you miss reviewing it? 12:42:40 I'm updating the job to build against the latest released glusterfs version, so at least we have nightly nfs-ganesha packages 12:42:51 kshlm: yes, me and shyam both 12:43:15 well, "fix the problem, not the blame" 12:43:35 kkeithley, Yup. 12:43:37 its not too bad, I do not think anyone except ganesha is affected, and it gives us a good reason to really fix it :D 12:43:49 ndevos: can other gfapi consumers be broken now? 12:43:57 ndevos: ah ok, you answered it already 12:44:04 hagarth: it's related upcall 12:44:07 jiffin: ok 12:44:17 hagarth: only if they use the cache-invalidation, and I think only Ganesha uses it 12:44:23 as long as we break things in master and quickly fix, we should be fine 12:44:26 ndevos: ok 12:44:38 Thanks for the update ndevos! 12:44:44 #topic Samba 12:45:08 Do we have any updates for Samba? 12:45:17 I don't see samba-folk around. 12:46:06 kshlm, Recently we did some re-factoring related to ACL handling code in vfs module for glusterfs in Samba. 12:46:44 some of the mdcache work happening should benefit samba performance afaik 12:46:52 and will be part of upcoming version 4.5 12:46:59 hagarth, Yes..yes..Definitely 12:47:25 4.5!! That's really late 12:47:39 lol 12:47:45 :D 12:47:54 Thanks for the update anoopcs 12:47:55 anoopcs: how is the richacl related work coming along? 12:48:30 hagarth, rjoseph is working on RichACL support for GlusterFS. 12:49:02 I think rjoseph is present here. 12:49:05 rjoseph: are you around? i see that listed on the roadmap for 3.9 12:49:10 He was during the roll call atleast. 12:49:36 sorry, was in discussion in samba channel 12:49:48 w.r.t Samba I don't think its upstream yet..patches have been sent by Andreas long time back. 12:50:15 Gluster related changes are going on, could not spend more time on this. 12:50:30 I will be taking this after the SSL changes I am doiing for libgfapi 12:50:43 we have some glfs_set_acl() functions I think, similar ones will be added for richacl 12:50:55 rjoseph, BTW I've got some info to share with you on that. 12:51:14 kshlm: regarding SSL or RichACL? 12:51:35 rjoseph: oh, are the ssl changes for libgfapi documented somewhere, a feature/design page perhaps? 12:51:49 SSL. About how reconnection failures are handled in the fuse client. 12:52:08 I'll catchup in #gluster-dev after the meeting. 12:52:11 No, currently investigating this as a bug, looks like I need a feature page for this after all 12:52:12 kshlm, ndevos : There are no functional changes w.r.t ACL coming with Samba 4.5..Just re-factoring to make it usable for both Ceph and GlusterFS. 12:52:41 sure kshlm 12:52:43 anoopcs: ok, good to know :) 12:52:55 rjoseph and anoopcs Thanks for all the information! 12:53:18 kshlm, np 12:53:22 #topic Community Infrastructure 12:53:33 misc, nigelb, Do you have anything to share? 12:54:58 Not much has happened here. 12:55:15 nigelb has been converting more Jenkins jobs to JJB. 12:55:43 He's also been fixing up the storage on netbsd guests. 12:56:22 and now we have more space on NetBD machines.. 12:56:30 *NetBSD 12:56:31 There was a small period of time yesterday, when www.gluster.org was opening blog.gluster.org 12:56:45 misc, fixed it quickly, and traced that down to a bug in Ansible. 12:56:57 anoopcs, Yes. We have more space now. 12:57:10 NetBSD regression should be more stable now. 12:57:21 And I that's it. 12:57:40 #topic Last weeks AIs 12:58:02 This is probably the first time all the AIs include the same person! 12:58:10 I'm on all of them. 12:58:20 oh, well done! 12:58:29 Ah, kshlm has said everything I wanted to. 12:58:37 Sorry, I was in the middle of something :) 12:58:40 4 AIs. 2 done. 12:58:53 ndevos did one and I did the other. 12:59:49 I'll be trying to get the centos-ci job for GD2 setup if I get time this week. 12:59:56 oh, sorry, didnt want to take that away from you... 13:00:43 ndevos, No worries. I greatfully accept any volunteering to do my AIs. 13:00:57 * anoopcs would like to inform that glusterfs-coreutils is also being run as part of CentOS CI jobs for Gluster 13:00:58 :) 13:01:05 anoopcs, Cool! 13:01:17 anoopcs: nice! 13:01:24 #link https://ci.centos.org/view/Gluster/job/gluster_coreutils/ 13:01:48 Thanks to all the help by ndevos 13:02:05 That's all about the AIs. 13:02:06 anoopcs: maybe we should add a badge to the README in the github repo? https://ci.centos.org/view/Gluster/job/gluster_coreutils/badge/ 13:02:24 I'll have my name on all the AIs for this week as well. 13:02:40 ndevos, It's already done by Craig. 13:02:50 oh, wow, he's fast! 13:02:55 https://github.com/gluster/glusterfs-coreutils 13:03:00 * kshlm notes that we're past scheduled time 13:03:33 We don't have any topics for discussion under Open Floor. 13:03:34 I'm done. 13:03:41 So I'll end the meeting if we're all done. 13:03:54 can i just suggest something? 13:04:07 paul98, Go ahead. 13:04:27 can someone look into the disk encryption as the docs are quiet poor, and google doesn't give you much info e.g i don't think many people have tried it, even though it's supported 13:05:07 the ssl side is spot on thanks to kshlm just disk encryption is a massive struggle 13:05:24 paul98: you mean encryption translator 13:05:26 paul98: ? 13:05:37 pranithk1, yup. the data-crypt xlator 13:05:46 pranithk1: yes 13:06:11 Sorry. It's just the crypt xlator. 13:06:36 Does anyone active have knowledge about this? 13:06:44 paul98: I think the guy who worked on it is not working on it anymore. I heard someone wants to work on it, may be prasanna. May be you should send me a mail? pkarampu@redhat.com 13:07:03 pranithk1, Let's have this on gluster-devel. 13:07:10 Don't do it in private. 13:07:14 paul98: I can connect you with someone interested. You could probably help us stabilize? 13:07:14 or gluster-users@gluster.org 13:07:26 it's something we want to use here 13:07:27 kshlm: that works better I guess 13:07:43 just the lack of documention doesn't help, 13:07:50 and it seems no one has tried it before 13:07:55 paul98: Well, we don't know what is the state of it. We can help you with it. 13:07:57 which i'm surprised at 13:08:11 but yes i'm keen to help out where needed 13:08:17 The spec in glusterfs-specs lists jdarcy and Edward Shishkin as the owners. 13:08:44 mainly it was Edward. (oops, fixed the blame) 13:08:46 kshlm: link? 13:08:55 Pretty much all Edward's code. I had written an earlier version, but it's now lost in the mists of time. 13:08:59 paul98: https://github.com/gluster/glusterfs/blob/master/tests/encryption/crypt.t sample test case for cryt xlator 13:09:24 it might be helpful 13:09:36 yesterday JoeJulian posted this comment here in #gluster-meeting. 13:09:39 [18:48:18] Since I cannot attend community meetings in the middle of the night, I'll just drop this here and hope it gets mentioned. 13:09:40 [18:48:53] There are a couple of documentation issues that are months old, one of which shows no evidence of even having been looked at. Can someone please take a look at https://github.com/gluster/glusterdocs/issues and make some progress? 13:09:49 ndevos, https://github.com/gluster/glusterfs-specs/blob/master/done/GlusterFS%203.5/Disk%20Encryption.md 13:10:14 kshlm: thanks, that was for paul98 of course :) 13:10:30 kkeithley, I saw some activity on that from ppai. 13:11:32 ppai should be looking into that. 13:11:44 We're 10 minutes over time. 13:11:50 I'll stop the meeting now. 13:12:01 ah cool, i'll go have a read! thnks lads, sorry for making you run over! :) 13:12:06 If you're attending any event/conference please add the event and yourselves to Gluster attendance of events: https://public.pad.fsfe.org/p/gluster-events 13:12:06 Put (even minor) interesting topics on https://public.pad.fsfe.org/p/gluster-weekly-news 13:12:39 I don't know if we're using the backports etherpad anymore (I know I've not looked at it) but anyways 13:12:50 Use the following etherpad for backport requests  https://public.pad.fsfe.org/p/gluster-backport-requests 13:12:51 lol 13:13:05 Thanks for attending today's meeting everyone! 13:13:09 #endmeeting