Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2014-04-09

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:11 awheeler joined #gluster-dev
01:18 bala joined #gluster-dev
01:34 bharata-rao joined #gluster-dev
02:43 badone joined #gluster-dev
03:17 kdhananjay joined #gluster-dev
03:23 shubhendu joined #gluster-dev
03:37 itisravi joined #gluster-dev
03:55 kdhananjay joined #gluster-dev
04:17 ndarshan joined #gluster-dev
04:26 deepakcs joined #gluster-dev
04:27 _ndevos joined #gluster-dev
04:29 raghu joined #gluster-dev
04:31 ppai joined #gluster-dev
04:45 kdhananjay joined #gluster-dev
05:09 atinm joined #gluster-dev
05:34 sahina joined #gluster-dev
05:36 bala1 joined #gluster-dev
05:42 lalatenduM joined #gluster-dev
05:44 kanagaraj joined #gluster-dev
05:54 spandit joined #gluster-dev
06:17 ppai joined #gluster-dev
06:20 aravindavk joined #gluster-dev
06:21 hagarth joined #gluster-dev
06:46 an joined #gluster-dev
07:24 ppai joined #gluster-dev
08:10 an joined #gluster-dev
08:14 kdhananjay joined #gluster-dev
08:23 bala1 joined #gluster-dev
08:37 hagarth joined #gluster-dev
08:53 lalatenduM deepakcs, in http://wiki.qemu.org/ChangeLog/Next search for gluster
08:53 glusterbot Title: ChangeLog/2.0 - QEMU (at wiki.qemu.org)
08:53 deepakcs lalatenduM, yes i did
08:55 deepakcs lalatenduM, that only says support for live snap merge.. but ur response to one of the list Q was to support snap itself
08:55 deepakcs lalatenduM, afaik qemu support snapshotting for netowrk disk from a long time
08:55 lalatenduM deepakcs, yes thats right, but with libgfapi you need qemu2.0
08:56 lalatenduM deepakcs, I also talked to shilpa about , she mentioned she will use qemu2.0 for test
08:56 deepakcs lalatenduM, and i am trying to understnad the reason for qemu 2.0. why is it needed unless u wanna use live merge
08:57 deepakcs lalatenduM, i am just sayng (rather confirming) that for basic snap to work qemu 2.0 is not needed.. an older qmeu also shud work, agree ?
08:57 lalatenduM deepakcs, qemu snpshot on fuse mount would work fine with older qemu bersion
08:57 lalatenduM s/bersion/version/
08:57 deepakcs lalatenduM, no, i am referring to libgfapi
08:58 deepakcs lalatenduM, I am trying to understand why u think qemu 2.0 is a must.. can u explain that pls
08:58 lalatenduM deepakcs, that I am not sure of
08:58 lalatenduM deepakcs, which version of qemu you are talking about
08:58 lalatenduM deepakcs, May be I will check the change log of qemu for that
08:59 deepakcs lalatenduM, say qmeu 1.3 ( i think thats where libgfapi support was first added)
09:00 lalatenduM deepakcs, yeah , but qemu 1.3 does not had support for libgfapi+qemu snapshot
09:00 bharata-rao deepakcs,
09:00 bharata-rao (qemu) info block
09:00 bharata-rao rootdisk: gluster://bharata/test/F17-test (raw)
09:00 lalatenduM deepakcs, hagarth would have more info
09:00 bharata-rao (qemu) snapshot_blkdev rootdisk gluster://bharata/test/F17-test-snap
09:00 lalatenduM deepakcs, hagarth ^^
09:00 bharata-rao [root@bharata ~]# qemu-img info gluster://bharata/test/F17-test-snap
09:00 bharata-rao image: gluster://bharata/test/F17-test-snap
09:00 bharata-rao file format: qcow2
09:00 bharata-rao virtual size: 10G (10737418240 bytes)
09:00 bharata-rao disk size: 193K
09:00 bharata-rao cluster_size: 65536
09:00 bharata-rao backing file: gluster://bharata/test/F17-test
09:00 bharata-rao backing file format: raw
09:00 bharata-rao Format specific information:
09:00 bharata-rao compat: 1.1
09:01 bharata-rao lazy refcounts: false
09:01 bharata-rao deepakcs, ^
09:01 lalatenduM bharata-rao, what is the qemu version ?
09:01 deepakcs bharata-rao, thanks
09:02 deepakcs lalatenduM, when u say "qemu 1.3 does not had support for libgfapi+qemu snapshot" - have u tested it or how are u so sure abt it ?
09:02 bharata-rao lalatenduM, mostly latest git - QEMU emulator version 1.7.50, Copyright (c) 2003-2008 Fabrice Bellard
09:03 deepakcs bharata-rao, but i think the same shud work since qemu 1.3 too
09:03 lalatenduM deepakcs, had discussion with sas and others who were testing qemu+gluster
09:03 bharata-rao deepakcs, I would think so, unless I have missed something
09:04 deepakcs lalatenduM, need to reverify, bcos for basic snapshot to work.. there weren't anything else added to qemu (unless as bharata-rao said we are missign something)
09:04 bharata-rao deepakcs, there were issues with libvirt which you are aware of and for which there were patches
09:04 deepakcs bharata-rao, I remember there was some discussion on having qemu to support gluster:// as part of the backign file .. was that added later than 1.3 ?
09:05 bharata-rao deepakcs, that was libvirt not qemu
09:05 lalatenduM deepakcs, please talk to sas, shilpa and hagarth and if you find out something concrete please reply to the original mail thread in gluster-users
09:06 deepakcs bharata-rao, libvirt never had support for network disks. but i am saying for backing file to support network disk format.. that must go into qemu, rite ?
09:06 lalatenduM deepakcs, I did not test qemu+kvm , so I have info from others, which can be wrong which it seems now
09:06 deepakcs "backing file: gluster://bharata/test/F17-test" <-- for this to work
09:06 deepakcs bharata-rao, ^^
09:07 bharata-rao deepakcs, as I said earlier, it has been working from the beginning, can't remember any separate change done for that in qemu
09:07 deepakcs lalatenduM, sure, np. BTW there is nothing wrong to correct here.. I am just verifying if we really need to pitch users to use qemu 2.0 always ? i think no, bcos if they just want basic snapshot to be taken and deleted (without all the fancy of merge, commit) i think qmeu > 1.3 shud work too
09:07 deepakcs bharata-rao, ok.. thanks
09:08 an joined #gluster-dev
09:08 lalatenduM deepakcs, if qemu 1.3 works well, that would be a relief for me, because I am having a bad time trying to rebuild qemu2.0 on CentOS6.5 , as the dependent packages are not available on CentOS/RHEL6.5
09:09 deepakcs lalatenduM, as i said, someone need to invest some time to validate it :) My hunch is that it shud work.. again disclaimer here.. basic snap create/delete shud work.. not all the otehr fancy stuff related to snaps
09:10 lalatenduM deepakcs, I am looking at you for the confirmation :), may be ndevos will help
09:12 deepakcs lalatenduM, i thot sas was the right one to confirm :)
09:12 lalatenduM ndevos, can you try libvirt 1.2.3  with cloudstack + default qemu in CentOS6.5
09:12 deepakcs lalatenduM, whats default qemu version in centos6.5 ?
09:12 deepakcs lalatenduM, also if u have a VM with that setup, we can try to validate it
09:12 ndevos lalatenduM: lets see what version I am running now...
09:14 lalatenduM deepakcs, I have another suggestion :), I think you can send the same info "qemu 1.3 with libvirt 1.2.3" should work for qemu snapshot + libgfapi, and somebody should test it from the community"
09:14 lalatenduM as a reply to the original mail thread
09:14 lalatenduM and may be it would inspire somebody to test it :) and we can definitely try Ur setup
09:14 deepakcs lalatenduM, sure i can
09:15 ndevos lalatenduM: I've got libvirt-0.10.2-29.el6.x86_64, where can I find a newer libvirt?
09:15 deepakcs lalatenduM, in this whole discussion i lost which thread was it in the list ;-)
09:15 deepakcs i will find it
09:18 lalatenduM ndevos, the mail sub was "RPMs for libvirt-1.2.3 for CentOS now available" in gluster-users ... the repo is http://download.gluster.org/pub/gluster/glusterfs/libvirt/CentOS/
09:18 glusterbot Title: Index of /pub/gluster/glusterfs/libvirt/CentOS (at download.gluster.org)
09:18 ndevos lalatenduM: and qemu-kvm-0.12.1.2-2.415.el6.x86_64
09:20 lalatenduM ndevos, I think we need qemu 1.3 , where the code drop for libgfapi happened , right deepakcs ^^
09:20 ndevos lalatenduM: well, I'm on rhel-6.5, qemu+libgfapi has been backported for that
09:22 ndevos lalatenduM: dammit, I also need an update for libvirt-python...
09:22 lalatenduM ndevos, thats good, How can  check which version of EL libvirt libgfapi got backported , is the change log public, I mean, is qemu-kvm-0.12.1.2-2 is maintained upstream?
09:23 ndevos lalatenduM: rpm -q --changelog qemu-kvm | grep -i gluster
09:23 lalatenduM ndevos, thanks will check
09:28 lalatenduM ndevos, interesting, I could successfully install libvirt 1.2.3 in CentOS6.5
09:29 lalatenduM ndevos, but in your case we getting the libvirt-python issue
09:30 ndevos lalatenduM: yeah, obviously cloudstack-agent depends on libvirt-python :-/
09:30 lalatenduM ndevos, good that we good it early :)
09:31 lalatenduM ndevos, will look in to the issue, but as of now have to rush to a meeting
09:31 ndevos lalatenduM: same here :)
09:31 lalatenduM ndevos, :)
09:34 bala1 joined #gluster-dev
09:37 hagarth joined #gluster-dev
09:50 lalatenduM joined #gluster-dev
09:59 an joined #gluster-dev
10:00 ppai joined #gluster-dev
10:02 deepakcs joined #gluster-dev
10:04 spandit joined #gluster-dev
10:10 kdhananjay joined #gluster-dev
10:19 aravindavk joined #gluster-dev
10:49 shyam joined #gluster-dev
10:58 an joined #gluster-dev
11:02 lalatenduM joined #gluster-dev
11:18 kdhananjay joined #gluster-dev
11:36 ndevos jclift: you can try https://access.redhat.com/labs/heartbleed/ - as long as you have a RH-login :)
11:36 glusterbot Title: Heartbleed Detector (at access.redhat.com)
11:40 ira_ joined #gluster-dev
11:49 deepakcs joined #gluster-dev
11:53 ndevos bfoster: you seem to be one of the FUSE/mount maintainers, could you review http://review.gluster.org/7278 this week?
11:53 glusterbot Title: Gerrit Code Review (at review.gluster.org)
12:07 kkeithley we're not running https: on build.gluster.org
12:08 kkeithley and that page is pretty dodgy. I'm signed in and I don't always get the applet
12:12 ndevos hmm, if no https is used, whats the reason jclift wants to update openssl?
12:13 kkeithley that's my point! ;-)
12:13 shubhendu joined #gluster-dev
12:13 ndevos :)
12:19 kkeithley download.gluster.org is vulnerable, and it is running https:
12:21 kkeithley I've updated openssl on download. Do we need new keys?
12:23 kkeithley and I restarted httpd
12:23 ndevos well, 'need' is such a big word, it is possible that someone received some leaked memory bits, those bits could have contained (parts of) the private key
12:24 ndevos I'm not sure if there would be any evidence in the httpd logs or something...
12:24 kkeithley hmm, that applet doesn't tell me that my home web server running freebsd 8.2 ppc is vulnerable, giving me a false sense of security
12:31 ndevos it's the 1...*e* version of openssl that has the bug, older versions dont have it
12:35 kkeithley okay, good, thanks
13:08 awheeler joined #gluster-dev
13:13 tdasilva joined #gluster-dev
13:15 itisravi joined #gluster-dev
13:19 [o__o] joined #gluster-dev
13:36 lalatenduM joined #gluster-dev
13:38 deepakcs joined #gluster-dev
13:50 an joined #gluster-dev
14:04 jclift kkeithley ndevos: Interesting applet.
14:05 jclift kkeithley ndevos: supercolony.gluster.org is reporting as still vulnerable
14:05 jclift kkeithley: You want to update that one too?
14:06 ndevos jclift: its 'value added services' when you buy your support at Red Hat!
14:06 jclift Oh, the access.redhat.com page that has extra info is really useful too.
14:06 jclift I didn't realise it's only the RHEL 6.5 shipped packages that are vulnerable.
14:07 ndevos and well, anything that contains a backport of the openssl *e* version *cough* RHS *cough*
14:07 jclift OpenSSH isn't affected either, which is good info.  That's why I was worried about build.gluster.org.
14:08 ndevos jclift: you worry too much, if you're not careful, your hair might go grey, or you'll gte bald
14:08 itisravi joined #gluster-dev
14:08 jclift ndevos: Soh!
14:08 jclift Doh!
14:09 wushudoin joined #gluster-dev
14:14 wushudoin joined #gluster-dev
14:24 jclift Well, I have access to download.gluster.org (supercolony-gen1 VM in rackspace).  That's the one kkeithley has updated, and is no longer vulnerable.
14:25 jclift We still need to update supercolony.gluster.org, which is running https and is vulnerable.  I don't seem to have ssh access to it.
14:25 jclift kkeithley: Do you?
14:36 johnmark w00t
14:36 johnmark reg for the storage hackathon is >110 people now
14:43 kkeithley is that review.gluster.org? No, I don't have access
14:51 johnmark jclift: I updated that one
14:51 johnmark jclift: I updated both of those, package-wise
14:51 jclift johnmark:
14:51 jclift johnmark: The supercolony.gluster.org one?
14:51 johnmark jclift: yes
14:52 johnmark jclift: I separate them into www.g.o and download.g.o
14:52 jclift johnmark: It's still showing as vulnerable.
14:52 johnmark jclift: what's showing as vulnerable?
14:52 * jclift tested it like 5 secs ago
14:52 johnmark jclift: how did you test?
14:52 jclift johnmark: https://access.redhat.com/labs/heartbleed/
14:52 glusterbot Title: Heartbleed Detector (at access.redhat.com)
14:52 johnmark huh... ok
14:52 * johnmark checks
14:52 jclift johnmark: There's an app there that connects to a box and checks if it's vulnerable
14:52 jclift johnmark: Did you restart http/https?
14:53 johnmark Oh!
14:53 jclift Heh
14:53 johnmark jclift: the mod_ssl package was updated when I logged in
14:53 johnmark but they have since updated the openssl packages
14:53 johnmark now updating :)
14:53 jclift :)
14:53 johnmark heh
14:53 johnmark thanks!
14:54 johnmark jclift: and you already did download.g.o?
14:54 jclift johnmark: kkeithley did
14:54 johnmark alright - I think we're set
14:54 jclift Yep.
14:54 jclift johnmark: Next thought... any idea how we can get extra storage allocated to build.gluster.org?
14:55 kkeithley GlusterFS Community Meeting in five minutes in #gluster-meeting @freenode
14:55 jclift johnmark: Or are we better off to just get it migrated to a new VM in Rackspace, where we'll be able to add be able to add new storage easily?
14:56 johnmark jclift: that would be my suggestion
14:57 jclift k
14:57 * jclift is running late updating the Agenda
15:01 jclift k, meeting time
15:02 lpabon joined #gluster-dev
15:07 jdarcy joined #gluster-dev
15:22 wushudoin joined #gluster-dev
15:40 hchiramm__ joined #gluster-dev
15:48 jclift That didn't feel like a very satisfying meeting. :/
15:49 purpleidea i blame the chair!
15:49 purpleidea (just kidding)
15:49 jclift Heh ;D
15:49 purpleidea what was unsatisfying to you?
15:49 purpleidea is d.g.o down again?
15:49 * kkeithley thinks there must be a joke about herding cats that applies
15:50 kkeithley no, it's not down
15:50 kkeithley I just signed on
15:50 purpleidea is there any reason why gluster-api package shouldn't be installed on all servers that are gluster hosts? what about clients? is it ever needed on the clients?
15:50 jclift We normally get a bunch of status answers, but today there wasn't much representation from hagarth/lala/kshlm/etc, who are normally there.
15:50 semiosis purpleidea: http://downforeveryoneorjustme.com/download.gluster.org
15:50 glusterbot Title: Down For Everyone Or Just Me -> Check if your website is down or up? (at downforeveryoneorjustme.com)
15:51 ndevos purpleidea: glusterfs-api is needed by anything that uses libgfapi, mostly those systems are clients
15:51 purpleidea ndevos: so not needed on clients?
15:52 ndevos purpleidea: qemu+libgfapi needs it, is a hypervisor a client?
15:52 hchiramm__ jclift, as I said, everyone is out for dinner ..
15:52 purpleidea ndevos: i mean, does a client that uses libgfapi need the api package?
15:52 purpleidea or is it just needed on the server?
15:52 ndevos purpleidea: also, samba is often co-located on storage servers, and the vfs_glusterfs plugin needs it, but it is possible to the the vfs_glusterfs plugin on non-storage servers
15:53 ndevos purpleidea: libgfapi *is* glusterfs-api
15:53 hchiramm__ afaict , client need api package as ndevos said
15:53 hchiramm__ if the client want to use libgfapi
15:53 purpleidea ndevos: so a puppet module should install glusterfs-api on each SERVER and on each CLIENT. right?
15:53 semiosis couldn't hurt
15:54 purpleidea semiosis: that's the answer i'm looking for :)
15:54 semiosis install all the things \o/
15:54 jclift hchiramm__: Yeah.  :/
15:54 purpleidea semiosis: yeah, i normally do, but then people complain. blah blah. deps
15:54 ndevos purpleidea: well, it should get pulled in automatically...
15:54 semiosis haters gonna hate
15:54 purpleidea ndevos: by what?
15:55 ndevos purpleidea: if you install qemu and it was linked against lingfapi, the rpm will have an automatically generated requires on the library, yum resolves that dependency to glusterfs-api and installs it
15:55 ndevos s/lingfapi/libgfapi/
15:55 purpleidea cool
15:56 ndevos purpleidea: that counts for sambe-vfs-gluster (or whatever the package is called) too
15:56 ndevos (definitely sambA-<something>)
15:57 hchiramm__ iic, glusterfs-libs pulls glusterfs-api as a dependency
15:58 * hchiramm__ need to confirm though
15:58 jobewan joined #gluster-dev
15:58 ndevos hchiramm__: that would be wrong
15:58 hchiramm__ thx for confirming :)
15:58 ndevos glusterfs-api depends on glusterfs-libs :)
15:58 purpleidea does the version of glusterfs-server RPM always match glusterfs-api RPM ?
15:59 ndevos it could be bug though, but glusterfs-api is not needed to be installed on a storage server
15:59 ndevos purpleidea: yes, it should, unless glusterfs-server is not installed, and glusterfs-api is
15:59 purpleidea ndevos: wait so it's now not required on the server?
15:59 semiosis purpleidea: nothing in glusterfs depends on the api
16:00 ndevos purpleidea: glusterfs-api is optional, depending on what components there are installed on the server
16:00 semiosis the api is for *other things* to access glusterfs
16:00 semiosis like samba
16:00 ndevos and QEMU
16:00 semiosis like i said, not required, but it couldn't hurt
16:01 ndevos purpleidea: if glusterfs-api is installed, it should have the same version as glusterfs-libs - and glusterfs-server depends on glusterfs-libs too (versions need to match)
16:04 an joined #gluster-dev
16:13 purpleidea for review: https://github.com/purpleidea/puppet-gluster/tree/feat/libgfapi
16:13 glusterbot Title: purpleidea/puppet-gluster at feat/libgfapi · GitHub (at github.com)
16:18 hagarth joined #gluster-dev
16:47 johnmark purpleidea: *sigh*
16:47 purpleidea johnmark: what's up?
16:47 purpleidea whaddya need?
16:47 purpleidea i bet it's a patch!
17:04 semiosis well shucks.  looks like the glusterfs MIR isn't going to make it in time for qemu support in trusty
17:12 purpleidea MIR?
17:14 semiosis Main Inclusion Request.  the procedure by which packages are elevated from the Universe repo to Main
17:40 lalatenduM joined #gluster-dev
17:51 shubhendu joined #gluster-dev
17:59 johnmark semiosis: :(
18:00 johnmark boo
18:29 kkeithley johnmark: when you said "man the booth" you were talking about the Gluster booth in the Developer Lounge, right?
18:39 _Bryan_ joined #gluster-dev
18:44 shyam joined #gluster-dev
18:48 johnmark kkeithley: yup!
20:40 lpabon joined #gluster-dev
20:54 tdasilva left #gluster-dev
22:13 hchiramm__ joined #gluster-dev
23:36 badone joined #gluster-dev
23:38 badone joined #gluster-dev
23:50 badone joined #gluster-dev

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary