12:01:21 #startmeeting 12:01:21 Meeting started Wed Oct 22 12:01:21 2014 UTC. The chair is davemc. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:01:21 Useful Commands: #action #agreed #halp #info #idea #link #topic. 12:01:40 who's here? 12:01:42 * kkeithley is here 12:01:49 * davemc is here 12:01:52 12:01:52 * ndevos is here 12:01:52 * JustinClift waves enthusiastically 12:01:56 * lalatenduM lurks 12:02:06 ;) 12:02:11 etherpad is https://public.pad.fsfe.org/p/gluster-community-meetings 12:02:34 #action FreeBSD maintainer 12:02:36 lalatenduM: That's lurking? ;) 12:02:42 any updates or confirms? 12:03:15 I sent a note off to his twitter handle, didn't hear back 12:03:23 jdarcy, :) 12:03:31 okay 12:03:52 #item new slave vm? 12:04:00 JustinClift, update? 12:04:05 * JustinClift looks 12:04:06 1 sec 12:04:11 davemc: not #item, but #topic :) 12:04:28 so noted. corrected on second try 12:04:38 Ahhh osrry, wasn't watching this channel 12:04:41 said I was groggy 12:04:47 So, FreeBSD maintainership is in progress 12:05:07 shocked I am, shocked, JustinClift 12:05:10 Craig Butler has initial port in upstream FreeBSD bug system, needing initial approval + import: 12:05:20 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194409 12:05:36 Once that's done, I think he's the official maintainer 12:05:42 k 12:05:44 * JustinClift isn't completely clued up on their processes tho 12:05:55 It seems like it could take a while 12:06:09 Upstream FreeBSD is a bit overloaded it seems 12:06:18 your other item, slave VM 12:06:18 ? 12:06:22 I nuked the VM 12:06:25 as per comments 12:06:42 I couldn't get the regression tests to pass in F20 with XFS, and didn't have time to investigate 12:06:43 FreeBSD was overloaded back in 2001 when I was working iSCSI issues 12:06:49 k 12:06:50 I'll pick i tup next week when back home 12:07:04 #item web site 12:07:09 * JustinClift has ignored it 12:07:12 updated in pad 12:07:32 but some progress. some changes have been made and pushed, waiting merge 12:07:54 * davemc doesn't have merge access 12:07:58 pushed to git or pushed to live site 12:08:08 so dependent on the kindness of strangers 12:08:11 to git 12:08:12 Ahhh. You need someone to merge them in git? 12:08:15 * JustinClift can do that 12:08:26 I can also give you access to commit stuff yourself too 12:08:29 I was _told_ not to merge to live site 12:08:32 * JustinClift will get that done shortly 12:08:35 Ahhh 12:08:50 actually, I was yelled at to not merge to live site 12:08:54 k, lets discuss ofline after meeting ;) 12:08:58 k 12:09:25 and please have the process of merging documented on the infra list :) 12:09:30 also, web site stuff. I'm identifying broken links so if anyone has a fav broken link, let me know 12:09:53 I never can find the link to the wiki on the website? 12:10:04 ndevos: Yeah, I have that problem too 12:10:08 Needs to be in the top menu 12:10:20 #action improve web site update process docs 12:10:42 moving on 12:10:44 #idea add a link to the wiki in the top mene 12:10:47 *menu 12:10:56 ignoring changes to nav bar yet again 12:11:06 12:11:21 #item Tiering/Bitrot 12:11:26 * ndevos has the wiki links in his history, so he'll find it 12:11:50 do we need to continue this here, move it gluster.next or table it? 12:12:14 I gotta go 12:12:21 I find the wiki by finding the blog 12:12:25 who's working on BitRot/Tiering *and* in this meeting? 12:12:27 see ya JustinClift 12:12:31 Full data classification is part of Gluster4 as currently planned. 12:12:32 * JustinClift waves 12:12:47 BitRot and tiering are subsets of that, which might appear earlier. 12:12:54 jdarcy, so move to gluster.next 12:13:30 I'm OK with it either way. Dan/Joseph are the ones currently working on it, and I don't see either here. 12:13:54 k. just don't want to lose it. was a great discussion here and email 12:14:07 #item 3.4 12:14:20 (is being updated on the pad as we speak) 12:15:00 any other updates on 3.4.6? 12:15:13 Raghavendra's DHT patches are merged 12:15:18 patch failed regression? 12:15:26 Pranith's AFR patch is failing regression atm 12:16:16 Pranith online? 12:16:25 looking at it? 12:16:49 k, moving on 12:16:56 #item 3.5 12:17:11 any updates? 12:17:31 many patches merged 12:17:37 one thing, I added pointers (from the wiki) to both 3.4beta and 3.5beta 12:17:47 planned 2nd beta for later this week 12:18:05 since it was unclear what the current state of things is 12:18:26 I'd like to add a new "info" box to the front page of gluster.org 12:18:39 calling out the current status of stuff 12:19:08 ndevos, tks 12:19:09 sure just let us know what details you need 12:19:17 will do 12:19:28 anything else? 12:19:38 Humble, lalatenduM: you can build packages for beta2 this week? 12:20:05 you guys have holidays tomorrow/friday, so maybe this week is bad 12:20:21 we can push a week out 12:20:22 I guess a beta2 early next week then 12:20:28 I can help with that 12:21:01 we'll plan it that way then 12:21:04 sure, but well, early next week should be ok too I think 12:21:14 either way is fine with me 12:21:31 you decide. I'm just here for the view 12:21:53 #item 3.6 12:22:19 hey, there's my topic :) 12:22:42 intend releasing beta4 this week and if all goes well, we can have 3.6 GA next week 12:22:43 quick question. I reviewed the Planning 3.6 doc for new features. Is it accurate? 12:22:50 davemc: yes, it is accurate 12:22:53 hagarth++ 12:23:13 we want to do a google hang out on the release 12:23:29 davemc: that would be cool 12:23:55 should we do a release readiness check next week in this meeting? 12:24:21 #action release readiness check for 3.6 on Oct 29 12:24:43 cool 12:25:02 I am also looking for a maintainer for release-3.6. If anybody is interested, please let me know. 12:25:10 date in website is fixed to 29-Oct now 12:25:38 (crickets) 12:25:42 maybe two maintainers? So it's not such a burdensome task? 12:26:04 ndevos, will try to do that (beta2) 12:26:15 kkeithley: sure, am all in favor of that 12:26:42 Will leave maintainer(s) question open for now 12:26:47 davemc: ok 12:26:59 * davemc is no longer technical enough to do this 12:27:25 are there any known blockers for release right now? 12:27:39 think I saw something SSH/Poodle related 12:27:55 ah, pad updated like magic 12:28:11 davemc: right, jdarcy & me will reach out to you on an advisory we would need to post on gluster.org 12:28:14 is there something we should squeeze into 3.4.6 for SSH/Poodle? 12:28:28 Probably. Just a sec. 12:28:44 http://review.gluster.org/#/c/8962/ 12:29:26 Also https://bugzilla.redhat.com/show_bug.cgi?id=1155328 12:29:26 jdarcy: do you know if there are backports for that yet? for 3.4, 3.5 and 3.6? 12:29:33 jdarcy, is that the text of wht we need post? 12:29:39 There are no backports yet. I only settled on a solution last night. 12:29:50 okay, thanks 12:30:00 davemc: I don't think that's the text, but it can be the seed info for some. 12:30:14 I'll try to craft some from that 12:30:23 please keep me in this loop 12:30:26 BTW, reviews would be even more appreciated that usual. 12:30:55 jdarcy: on my list for this week 12:31:01 #action davemc to draft advisory text from http://review.gluster.org/#/c/8962/ 12:31:34 The workaround is inelegant (especially as the list in future versions of OpenSSH might change) but AFAIK effective. 12:32:07 jdarcy: the solution looks simple enough, but I'm a complete noob regarding those ciphers 12:33:04 okay, any more 12:33:28 jdarcy: really OpenSSH, or did you mean OpenSSL? 12:33:37 I meant OpenSSL. 12:33:41 if the 1st, I really need to look into it more 12:33:47 ah, ok, pfew 12:34:03 #item downstream 3.6 release issue 12:34:12 pad has short update 12:34:31 any commentary, kkeithley ? 12:34:32 yup 12:34:54 just need to wrap it up and deploy I think. ndevos, agree? 12:34:56 is that the RHEL-6.6 now provides a glusterfs-3.6 package issue? 12:35:17 that's a downstream issue that Red Hat releng is handling 12:35:39 I dont know how Red Hat releng handles it? 12:35:46 we're talking about our gluster-release RPM, without the YUM priority package 12:36:00 yeah, that plan looks good to me 12:36:15 I have not checked the latest version of the .spec though 12:36:33 Should I add the POODLE bug to blockers for 3.6? 12:36:39 I haven't changed it since our last conversation. I need to remove the Requires yum-priority-plugin 12:36:41 jdarcy, yes 12:36:45 or yum-plugin-priority 12:37:09 I am against yum-plugin-priority 12:37:27 lalatenduM: yes, we are all agreed on that I think 12:37:36 kkeithley, ok 12:37:37 jdarcy: and probably 3.4 and 3.5 too? 12:38:05 agreed from me 12:38:41 lalatenduM: yum-plugin-priority is the recommendation that we give, but we will not enforce it on our users 12:39:07 somehow we need to tell people to review their system and explicitly install yum-plugin-priority if they want 12:39:32 but not automatically install it as part of installing the gluster-release RPM. 12:39:44 ndevos, I would prefer yum update -y --exclude=glusterfs\*-3.6.0.29-2.el6 12:40:23 lalatenduM: we can not expect all users to execute that each time when they want to 'yum update' 12:40:43 installing yum-plugin-priority is ok, but not marking repos with priority 12:41:13 ndevos, I understand 12:41:33 versus marking repo with priority but not installing yum-plugin-priority? 12:41:38 but once you give a repo some priority , it changes the behavior of yum 12:41:42 we do not know if a user has a repository with priority, before we pull in the plugin - that would alter their existing config in unexpected ways 12:42:16 but what if they already have yum-plugin-priority already installed? 12:42:22 let's take this off-line? 12:42:24 * ndevos prefers to mark our repository with a priority 12:42:33 to #gluster-devel that is 12:42:36 kkeithley, agree 12:42:40 if they have the plugin installed, all should be fine :) 12:42:43 anything more on the RHEL 6.6. I have it as underway, not fixed yet. off line discussion 12:42:43 kkeithley, +1 12:43:06 kkeithley: you want to send an email with the proposal? 12:43:10 sure 12:43:15 cool, thanks 12:43:25 #action kkeithley ndevos to update us on RHEL-6.6 issue 12:43:42 #item gluster.next 12:44:12 any more discussion on small file stuff? 12:44:32 who added that topic? 12:45:02 which topic? small file or overall next plan? 12:45:06 Just updated the Etherpad a bit. 12:45:06 the small file one 12:45:48 there was a pretty active discussion last week, I think 12:45:51 hagarth and I talked about a (semi-)formal kickoff maybe next week 12:45:56 would have to check log 12:46:23 davemc: ok, let us follow up small files next week. I will also build some context from last week's logs. 12:46:35 hagarth, okay 12:46:50 we'll re-open it next week 12:47:04 right, as we taper off 3.6, we can simultaneously plan out & execute both 3.7, 4.0 12:47:20 The link Krutika gave last week is for a way-out-of-date plan. 12:47:40 Being the new guy, not sure how we plan the next stuff 12:47:50 "Plan" might not be the word. ;) 12:48:20 and I want to make sure we try to get a broader community ivolvement 12:48:27 s/iv/inv/ 12:48:39 davemc: planning is a polite term for unleashing chaos ;) 12:48:51 Technically, there are supposed to be real feature pages and then community discussion of those. 12:49:08 okay, a potential broad target of stuff that might happen or not 12:49:24 In reality, some of those pages are sketchy, some agreed-to items never get done (which is OK), and half of what does go in was never part of the planning process. 12:49:41 jdarcy, I assume this is on the wiki 12:50:06 davemc: Not sure about that, except maybe as part of the per-release planning pages. 12:50:35 davemc: I think we have an informal writeup somewhere, will try to look it up after the meeting 12:50:38 * davemc really needs to dig deeper into how we figure what comes next-ish 12:50:47 hagarth, tks 12:50:55 Personally, I would *really* like to see an actual voting system on next-release features. 12:51:06 ah, that reminds me, people in #gluster tend to be interested in a feature like 'spare brick', similar to a hot-standby-disk for a raid-config - is that a feature request we have somewhere already? 12:51:33 ndevos: yes 12:51:35 ndevos: Don't recall seeing one. I could write one up (perhaps with your help) if you want. 12:51:38 and yes, some voting would be amazing 12:52:08 we will be releasing a short survey on the state of GlusterFS this week. 12:52:10 I also plan to seek feedback from the broader community on what they would like to see in 3.7 & beyond 12:52:35 Once we get some concrete ideas, we can set up polling / voting 12:52:50 davemc: cool 12:53:14 hagarth: uhm, you agree that it gets asked on regular basis, or that there is a feature request documented somewhere already? 12:53:19 "Polling" is probably a better word. 12:53:31 hagarth, would love to hear more on how to get the community involved. 12:53:40 I prefer polling 12:54:11 ndevos: I think there's a feature page or an enhancement request 12:54:34 davemc: if there are any Gluster Community days scheduled, we should probably get the people attending there to do a vote 12:54:38 well, there kind of is 12:54:44 hagarth: ah, cool 12:54:59 ndevos, one comin up in Paris in a couple of weeks. 12:55:14 but would need a target list 12:55:33 davemc: oh, we do? maybe we should promote it a little more? 12:55:45 davemc: happy to help with the questionnaire for that event 12:55:45 http://www.gluster.org/community/documentation/index.php/Features 12:56:01 ndevos, well, was scheduled befor emy arrival, but we're trying to make it happen 12:56:30 I've got another meeting in 5 minutes, sorry 12:56:47 davemc: cool, but either way, I've not heard of its confirmation before - and Paris is quite good reachable for me 12:56:58 jdarcy: probably a bz request, will try to look it up 12:57:13 will send you what i have ndevos after meeting. Looking for speaker... 12:57:27 davemc: sure 12:57:30 ndevos, do I have your email? 12:58:07 ndevos: https://www.eventbrite.com/e/inscription-gluster-cloud-night-paris-10413255327 12:58:32 Unless there are objections, I'd like to table the other item on community involvement till next week? 12:58:32 davemc: just add @redhat.com to my nick 12:58:41 ndevos, got it 12:59:05 * davemc needs caffeine and an IRC-email translator 12:59:06 hagarth: thanks, we should really do more blog posts and use the announce list for things like that 12:59:07 davemc: sounds good to me 12:59:32 ndevos, hope I can get ahead of this. Was news to me as well 12:59:45 davemc: I'm sure you can :) 12:59:57 ndevos: yeah, we just noticed that the event was planned and we had people already signing up for the event ;) 13:00:12 hagarth: oh, people are attending! 13:00:27 ndevos: quite a few signups, yes 13:00:55 any more today folks 13:00:59 ? 13:01:01 hagarth: very cool! 13:01:15 davemc: thanks! 13:01:38 ndevos, I may even try to make an appearance, if I can plan more than one day in Europe 13:01:53 okay, closing meeting 13:01:57 and thanks 13:02:08 #endmeeting