Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2014-09-23

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

All times shown according to UTC.

Time Nick Message
01:17 aviksil joined #gluster-dev
01:24 krishnan_p joined #gluster-dev
03:56 itisravi joined #gluster-dev
03:57 kanagaraj joined #gluster-dev
04:05 bharata-rao joined #gluster-dev
04:06 ndarshan joined #gluster-dev
04:09 aravindavk joined #gluster-dev
04:09 nishanth joined #gluster-dev
04:17 shubhendu joined #gluster-dev
04:19 aviksil joined #gluster-dev
04:26 atinmu joined #gluster-dev
04:36 anoopcs joined #gluster-dev
04:37 lalatenduM joined #gluster-dev
04:43 rafi joined #gluster-dev
04:45 soumya_ joined #gluster-dev
05:04 spandit joined #gluster-dev
05:07 aravindavk joined #gluster-dev
05:16 aviksil joined #gluster-dev
05:16 deepakcs joined #gluster-dev
05:17 aravindavk joined #gluster-dev
05:32 raghu joined #gluster-dev
05:34 kdhananjay joined #gluster-dev
05:38 hagarth joined #gluster-dev
05:39 jiffin joined #gluster-dev
05:42 ppai joined #gluster-dev
05:43 pranithk joined #gluster-dev
05:43 Gaurav_ joined #gluster-dev
05:56 atalur joined #gluster-dev
05:56 bala joined #gluster-dev
05:56 soumya_ joined #gluster-dev
06:05 soumya_ joined #gluster-dev
06:07 pranithk joined #gluster-dev
06:10 bala joined #gluster-dev
06:15 atalur joined #gluster-dev
06:19 hagarth joined #gluster-dev
06:34 kshlm joined #gluster-dev
06:39 deepakcs joined #gluster-dev
06:49 atinmu joined #gluster-dev
07:05 deepakcs joined #gluster-dev
07:23 pranithk joined #gluster-dev
07:25 kaushal_ joined #gluster-dev
07:26 kaushal_ joined #gluster-dev
07:33 RaSTar joined #gluster-dev
07:40 hagarth joined #gluster-dev
07:40 atinmu joined #gluster-dev
07:48 ramon_dl joined #gluster-dev
08:08 aravinda_ joined #gluster-dev
08:29 kanagaraj joined #gluster-dev
08:50 soumya joined #gluster-dev
08:54 soumya joined #gluster-dev
09:01 kanagaraj joined #gluster-dev
09:04 vimal joined #gluster-dev
09:08 bala joined #gluster-dev
09:20 shubhendu joined #gluster-dev
09:22 ndarshan joined #gluster-dev
09:59 Gaurav__ joined #gluster-dev
10:04 ws2k333 joined #gluster-dev
10:20 soumya joined #gluster-dev
10:20 shubhendu joined #gluster-dev
10:20 atinmu joined #gluster-dev
10:20 hagarth joined #gluster-dev
10:21 ndarshan joined #gluster-dev
10:23 bala joined #gluster-dev
10:35 pranithk JustinCl1ft: I wonder why http://build.gluster.org/jo​b/rackspace-regression-2GB/ is not triggering build #694 even though nothing else is running... any clue?
10:37 ppai joined #gluster-dev
10:43 ndevos pranithk: not sure if you solved it already, but there seem to be a lot of tests running atm?
10:43 pranithk ndevos: yeah, that is the reason I think :-)
10:44 ndevos okay, so your "even though nothing else is running..." threw me off :)
10:45 kkeithley1 joined #gluster-dev
10:54 shyam joined #gluster-dev
11:04 atinmu joined #gluster-dev
11:12 edward1 joined #gluster-dev
11:19 ndarshan joined #gluster-dev
11:22 nishanth joined #gluster-dev
11:28 kanagaraj joined #gluster-dev
11:28 soumya joined #gluster-dev
11:33 ppai joined #gluster-dev
11:39 hagarth joined #gluster-dev
11:41 pranithk left #gluster-dev
11:55 soumya__ joined #gluster-dev
12:00 ndevos REMINDER: Gluster Bug Triage meeting starting in a bit in #gluster-meeting
12:10 ppai joined #gluster-dev
12:18 itisravi joined #gluster-dev
12:38 hagarth joined #gluster-dev
12:46 kanagaraj joined #gluster-dev
13:03 cristov joined #gluster-dev
13:03 soumya__ joined #gluster-dev
13:04 cristov joined #gluster-dev
13:13 shyam joined #gluster-dev
13:46 deepakcs joined #gluster-dev
13:53 lalatenduM joined #gluster-dev
14:09 bala joined #gluster-dev
14:13 wushudoin| joined #gluster-dev
14:22 tdasilva joined #gluster-dev
14:38 hagarth joined #gluster-dev
15:04 jobewan joined #gluster-dev
15:09 kshlm joined #gluster-dev
15:12 hagarth joined #gluster-dev
15:43 hagarth joined #gluster-dev
15:47 dlambrig_ joined #gluster-dev
15:49 soumya__ joined #gluster-dev
15:53 Gaurav__ joined #gluster-dev
16:06 tdasilva joined #gluster-dev
17:12 tdasilva joined #gluster-dev
17:50 anoopcs joined #gluster-dev
17:50 hagarth1 joined #gluster-dev
18:10 kkeithley_ ndevos: ping. 3.6.0beta1 glusterfs-server rpms are getting     error: unpacking of archive failed on file /var/lib/glusterd/hooks/1/add-brick/post/d​isabled-quota-root-xattr-heal.sh;5421b216: cpio: open
18:10 ndevos kkeithley_: uh...
18:10 kkeithley_ but I can rpm2cpio the glusterfs-server rpm and cpio -tv or cpio -di without error
18:10 kkeithley_ any ideas why?
18:10 ndevos kkeithley_: what rpm is that?
18:11 kkeithley_ 3.6.0beta1 from download.gluster.org. el7 and f20 at least. I haven't tried others
18:11 ndevos kkeithley_: could it be that the rpm does not contain the directory /var/lib/glusterd/hooks/1/add-brick/post ?
18:12 ndevos kkeithley_: "rpm -qlp $RPM" should show that...
18:12 kkeithley_ %dir %attr(0755,-,-) %{_sharedstatedir}/glusterd/hooks/1/add-brick/post
18:13 ndevos hmm...
18:13 kkeithley_ /var/lib/glusterd/hooks/1/add-brick
18:13 kkeithley_ /var/lib/glusterd/hooks/1/add-brick/post
18:13 kkeithley_ /var/lib/glusterd/hooks/1/add-brick/p​ost/disabled-quota-root-xattr-heal.sh
18:14 ndevos kkeithley_: oh, and /var/lib/glusterd/hooks/1/add-brick/post does not exist yet, like, it is not a file or something?
18:14 ndevos or is it an issue that happens everywhere?
18:14 * ndevos is in a train, and may well loose connection soon
18:16 kkeithley_ that's a  mkdir -p %{buildroot}%{_sharedstatedir}/​glusterd/hooks/1/add-brick/post   in the spec
18:16 ndevos should be good, but the path should not exist on the client yet (as a non-directory)
18:16 kkeithley_ oh, maybe because there's a %ghost for it
18:16 ndevos oh, it should not be %ghost'ed
18:17 kkeithley_ yeah, new scripts, old %ghost.
18:17 ndevos that should be it, a %ghost is not included in the rpm, but a manual cpio will create the missing directories
18:18 kkeithley_ and no %ghost $parentdir either?
18:18 JustinCl1ft 3.6.0beta1-2 coming up... ?
18:18 kkeithley_ 3.6.0-0.2beta1, yeah
18:19 ndevos :-/
18:20 kkeithley_ need to edumacate our devs when they add a hook script to remove the %ghost
18:20 ndevos and the reviewers need to pay more attention to that...
18:21 kkeithley_ and our packagers should try installing the RPMs before they put them on d.g.o
18:21 kkeithley_ mistakes all around
18:21 ndevos uhm, yes, that too
18:21 kkeithley_ life goes on
18:22 ndevos and the package building too
18:22 ndevos I'm in a tunnel in a bit, cya tomorrow!
18:22 kkeithley_ ttyl
18:39 lalatenduM joined #gluster-dev
19:00 ramon_dl joined #gluster-dev
19:53 rafi1 joined #gluster-dev
20:27 lalatenduM kkeithley, regarding 3.6beta1 specfile
20:50 dlambrig_ left #gluster-dev
20:51 kkeithley_ lalatenduM: yes?
20:54 lalatenduM kkeithley, I have sent a mail abt it
20:54 lalatenduM there are some more issues
20:55 kkeithley_ one thing.... I believe we bumped the SO_VERSION of libgfapi.so
20:56 kkeithley_ we'll need to announce that on the fedora devel list. That's almost certainly what's breaking qemu, samba, (and ganesha)
20:57 lalatenduM kkeithley, yup, that is qemu is complaining
20:57 kkeithley_ I have a spec file that solves the -server install problem. I was just about to email it to you when I saw your ping
20:57 glusterbot kkeithley_: Please don't naked ping. http://blogs.gnome.org/mark​mc/2014/02/20/naked-pings/
20:58 lalatenduM kkeithley, awesome, I was scratching my head over it
20:58 lalatenduM kkeithley, what was the issue
20:58 kkeithley_ damn %ghost cruft
20:59 lalatenduM kkeithley, I thought we have cleaned them
20:59 kkeithley_ they are a twisty maze
21:00 lalatenduM kkeithley, oohno, there are so many of them :(
21:00 kkeithley_ but gluster isn't connecting to glusterd
21:01 JustinCl1ft kkeithley_: "and our packagers should try installing the RPMs before they put them on d.g.o"
21:01 lalatenduM kkeithley, ?? didn't get u
21:01 JustinCl1ft kkeithley_: Yeah, that's a checklist thing
21:01 lalatenduM JustinCl1ft, we do that most of the time
21:02 JustinCl1ft Recently created one for a wide project, after also forgetting a basic step: https://github.com/sqlitebrowser/s​qlitebrowser/wiki/Release-process
21:02 JustinCl1ft It's now become pretty comprehensive
21:02 kkeithley_ Oops I left a couple commented out %ghost stanzas
21:02 JustinCl1ft lalatenduM: Yeah.  Checklists are exactly for fixing "most of the time" problems :D
21:02 lalatenduM JustinCl1ft, :)
21:02 JustinCl1ft s/wide project/side project/
21:03 kkeithley_ most of the time surgeons don't leave instruments in the patients too. ;-)
21:03 lalatenduM lol
21:03 JustinCl1ft Another case in point (sic) :D
21:03 lalatenduM kkeithley, I think we should commit the the latest spec file changes to fedora dist git master
21:03 kkeithley_ why the change of heart?
21:04 lalatenduM kkeithley, thanks to Fedora updates policy document
21:04 lalatenduM :)
21:04 lalatenduM kkeithley, also as you said "<kkeithley_> we'll need to announce that on the fedora devel list. That's almost certainly what's breaking qemu, samba, (and ganesha)"
21:05 lalatenduM so we need the latest in master
21:05 kkeithley_ you lost me?
21:05 kkeithley_ libgfapi.so hasn't had a SO_VERSION before now. Now it does IIRC
21:06 lalatenduM kkeithley, right, so now qemu and samba need to fix their specfile and build top of the latest gluster i.e. 3.6 for rawhide right
21:06 kkeithley_ and libvirt is linked against it. So when we bump it they need to rebuild
21:07 lalatenduM yeah so we need to do fedpkg build for rawhide
21:07 lalatenduM and we need to commit our changes in rawhide/master branch
21:07 lalatenduM atleast
21:07 kkeithley_ when we ship 3.6.0 for real in Fedora
21:08 lalatenduM kkeithley, in rawhide ,yes
21:08 kkeithley_ I didn't think you wanted to commit the beta .spec to dist-git?
21:08 kkeithley_ and any other branch we ship it in. But we decided only in f22/rawhide. And maybe f21? f21 is still only alpha
21:09 lalatenduM kkeithley, change of heart
21:09 kkeithley_ we might get it in to f21? Or do you not want to do that?
21:09 kkeithley_ you had a change of heart?
21:09 lalatenduM kkeithley, post alpha we should not change the major version
21:09 lalatenduM kkeithley, yeah
21:10 kkeithley_ sure, that's fine with me
21:10 lalatenduM kkeithley, do u think we should ship 3.6 in f21?
21:10 kkeithley_ should? maybe. Could? certainly.
21:12 lalatenduM kkeithley, do u know when f21's GA?
21:12 kkeithley_ I don't
21:13 kkeithley_ I'm leaning further and further to ship one "major" in a particular Fedora release and don't go beyond. I know Niels feels that way.
21:14 kkeithley_ We should just agree that that's what we're going to do from now on. Our repos on download.g.o will allow people to jump forward to newer versions if they want them.
21:14 lalatenduM kkeithley, yes, I too agree
21:14 lalatenduM kkeithley, may be we should put this in some gluster wiki page
21:14 kkeithley_ weirdness. With the epel7 rpms, glusterd complains that /usr/lib64/glusterfs/3.6.0b​eta1/rpc-transport/rdma.so isn't there.   But it is
21:15 kkeithley_ and glusterd has a listener on port 27007, but gluster can't connect
21:15 lalatenduM kkeithley, iptable or selinux ? stopping it
21:16 kkeithley_ This is the longevity cluster. I had 3.5.2 running on it until I started trying to update to 3.6.0beta1 and ran into all the problems. I have iptables and selinux disabled
21:17 kkeithley_ not that I'm even using rdma
21:18 kkeithley_ glusterd listening on 24007 that is
21:18 lalatenduM kkeithley, hmm sounds like a bug
21:18 kkeithley_ but it's back to enforcing somehow.
21:19 kkeithley_ mmmph.
21:19 kkeithley_ how was it working before.
21:20 lalatenduM kkeithley, ohh
21:22 kkeithley_ that's better
21:25 kkeithley_ updating longevity cluster to 3.6.0beta1 was supposed to be quick-and-easy
21:25 systemonkey joined #gluster-dev
21:26 lalatenduM kkeithley, what the is use of %ghost
21:26 lalatenduM I am trying to find some documentation in internet , did not find any
21:27 kkeithley_ the whole %ghost thing is for declaring ownership of files and dirs. files and dirs that aren't installed by the rpm itself
21:28 kkeithley_ when when someone does `rpm -q --whatprovides /var/lib/glu.../hooks/1/add-brick/pre` then rpm can tell the user it belongs to glusterfs-server
21:28 lalatenduM kkeithley, cool
21:28 kkeithley_ so we have this convoluted logic in the spec file to mkdir a bunch of directories in .../rpmbuild/BUILDROOT/...
21:29 kkeithley_ which don't end up in the rpm file
21:29 kkeithley_ but now in 3.6 we've actually added some real file, so we don't need as many %ghosts
21:30 kkeithley_ some real files in the hooks
21:30 lalatenduM kkeithley, yeah all the dirs in hooks  does not need ghost I believe as we are creating them
21:30 lalatenduM in the spec file
21:31 kkeithley_ yes, we were always creating them. There's a bug in rpm/rpmbuild that requires we create them in order for %ghost to work. But when the rpm installs (cpio actually) it doesn't create the dirs.
21:32 kkeithley_ it's been a real pain to get rigth
21:32 kkeithley_ right
21:34 kkeithley_ in large part because it's not well documented. As you noted
21:34 kkeithley_ ;-)
21:34 kkeithley_ redhat
21:35 lalatenduM kkeithley, :),
21:36 lalatenduM kkeithley, ttyl, need to catch some sleep now
23:04 MacWinner joined #gluster-dev
23:40 bala joined #gluster-dev

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