Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2014-01-03

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

All times shown according to UTC.

Time Nick Message
01:51 aravindavk joined #gluster-dev
01:55 johnmark joined #gluster-dev
02:32 kdhananjay joined #gluster-dev
02:33 bharata-rao joined #gluster-dev
03:00 kshlm joined #gluster-dev
03:04 kdhananjay joined #gluster-dev
03:28 kdhananjay joined #gluster-dev
03:31 shubhendu joined #gluster-dev
03:45 itisravi joined #gluster-dev
03:54 ababu joined #gluster-dev
04:03 mohankumar__ joined #gluster-dev
04:14 aravindavk joined #gluster-dev
04:26 kdhananjay joined #gluster-dev
04:48 ndarshan joined #gluster-dev
04:48 ppai joined #gluster-dev
04:55 badone joined #gluster-dev
05:09 mohankumar__ joined #gluster-dev
05:43 kaushal_ joined #gluster-dev
05:59 mohankumar__ joined #gluster-dev
06:18 lalatenduM joined #gluster-dev
06:43 kaushal_ joined #gluster-dev
06:47 kdhananjay joined #gluster-dev
07:13 bala joined #gluster-dev
07:38 kdhananjay joined #gluster-dev
07:39 shubhendu joined #gluster-dev
08:19 xavih johnmark: I talked with pk and hagarth about the status of the disperse translator and the new modules
08:20 xavih johnmark: They are taking a first look at it
08:20 xavih lalatenduM: ping
08:20 lalatenduM xavih, pong
08:23 xavih lalatenduM: It's about the review of http://review.gluster.org/#/c/6626/4
08:23 xavih I'm not sure if there are strict rules about formating, and I'm not sure if it's important or not
08:23 xavih it's only a matter of formatting
08:24 xavih for what I see in other modules, labels are always at first column, the rest of code should be normally indented
08:25 lalatenduM xavih, yeah ,actually the formatting of this modules is not good
08:25 xavih however I'm not a core developer, and maybe this is unnecessary
08:26 xavih I can +1 this patch and wait for someone else to answer...
08:26 lalatenduM xavih, that does not matter :) neither I am . we can try to make better
08:27 lalatenduM xavih, please put your comments , so that I can work on them if possible
08:27 lalatenduM xavih, I am just trying to fix memory issues reported by Coverity
08:28 xavih lalatenduM: yes, I know. It's for that reason that I'm not sure if being so strict is good :P
08:29 kdhananjay joined #gluster-dev
08:29 xavih lalatenduM: I'll +1 but I'll also add some comments about formatting
08:29 xavih lalatenduM: is that ok ?
08:29 lalatenduM xavih, yup, should be fine , thanks for reviewing the patch :)
08:29 xavih lalatenduM: you're welcome :)
08:29 lalatenduM xavih, :)
08:34 lalatenduM xavih, I will put the lebel at 1st column and fix the formatting and send another patch
08:41 xavih lalatenduM: ok :)
08:46 xavih lalatenduM: Now I don't have any further objection :)
08:47 lalatenduM xavih, cool :) , thanks
09:06 hagarth massive failures in smoke tests
09:06 hagarth due to a regression run which ran for > 9 hours!
09:07 hagarth have re-triggered all the smoke tests that failed
09:25 ababu joined #gluster-dev
09:34 ppai joined #gluster-dev
09:50 ndarshan joined #gluster-dev
10:41 hagarth ndevos: ping
10:42 ndevos pong hagarth
10:42 hagarth ndevos: planning to get 3.4.2 out today. just posted release notes for review.
10:43 ndevos hagarth: oh, awesome!
10:44 hagarth will be triggering the release script shortly.
10:48 hagarth can anybody review the release notes for 3.4.2 - http://review.gluster.org/#/c/6639/ ?
10:49 ndevos hagarth: anything specific to pay attention to?
10:49 hagarth mostly formatted output from git log --pretty=oneline ;)
10:49 hagarth ndevos: nothing significant, a sanity check should do.
10:52 ndevos hagarth: any idea why http://build.gluster.org/job/smoke/5811/ gets aborted, or rather, how to fix that?
10:52 hagarth ndevos: I think there was a re-submission of the same patch
10:52 hagarth when the smoke tests were being run
10:52 hagarth hence jenkins decided to abort the run and queue a new smoke job.
10:53 ndevos hagarth: ah, that could be, thanks
10:55 ndevos hagarth: I'm happy with those notes :)
10:55 hagarth ndevos: thanks!
10:56 ndevos np!
11:00 pk joined #gluster-dev
11:04 hagarth okay, have triggered the release script for 3.4.2.
11:23 lalatenduM joined #gluster-dev
11:24 ndarshan joined #gluster-dev
11:25 hagarth and 3.4.2 is out!
11:26 pk hagarth: yay!
11:28 ndevos hagarth: cool, and tagged in the repo too?
11:29 ndevos ah, yes it is :)
11:29 hagarth ndevos: yeah
11:55 kdhananjay joined #gluster-dev
11:57 kkeithley joined #gluster-dev
12:01 itisravi_ joined #gluster-dev
12:16 shubhendu joined #gluster-dev
12:38 ira joined #gluster-dev
12:51 bala joined #gluster-dev
13:18 kdhananjay joined #gluster-dev
14:09 pk left #gluster-dev
14:18 hagarth1 joined #gluster-dev
14:49 jobewan joined #gluster-dev
15:15 wushudoin joined #gluster-dev
15:16 johnmark hagarth1: how are you feeling?
15:40 ira joined #gluster-dev
16:05 hagarth johnmark: thanks, not completely out of the woods yet
16:08 johnmark hagarth: bummer :(
16:09 johnmark ok, just found this from jperkin, who's at Joyent
16:09 johnmark 07:56 <jperkin> ok, most recent failure for glusterfs is here:
16:09 johnmark http://us-east.manta.joyent.com/pkgsrc​/public/reports/upstream-trunk32/20131​231.2033/glusterfs-3.4.1nb1/build.log
16:09 johnmark 07:57 <jperkin> can fix those reasonably easy in pkgsrc with some -Du_int*=uint*, but
16:09 johnmark would be nice to have it fixed upstream
16:09 johnmark 08:06 <johnmark> ah, interesting
16:09 johnmark 08:06  * johnmark looks
16:09 johnmark 08:07 <johnmark> could you do me a favor? Try that with the *just* released 3.4.2 - and
16:09 johnmark report the results to gluster-devel?
16:09 johnmark 08:08 <johnmark> so it's a matter of making the compile-time vars be more portable?
16:09 johnmark 08:08 <jperkin> uint_* are generally a lot more portable than u_int*
16:09 johnmark hagarth, a2_ : ^^^
16:09 johnmark 08:09 <jperkin> as they are C99/SUSv3
16:09 johnmark 08:09 <johnmark> got it, ok
16:13 hagarth johnmark: ok. where does jperkin hang out if we have more questions?
16:14 johnmark hagarth: he's in #illumos and probably #smartos
16:14 johnmark confirmed - also in #smartos
16:14 johnmark ira: ^^^
16:16 ira Yep.
16:16 ira I'd use #illumos.
16:16 hagarth johnmark, ira: ok, thanks
16:16 ira There may be a pkgsrc channel of some form.
16:16 ira if so use that.
16:17 johnmark ira: indeed there is - #pkgsrc
16:17 johnmark and many fewer people, thus probably better :)
16:17 ira :)
16:18 ira Are we going to want an illumos build box long term?
16:19 johnmark ira: I think so. I wonder if I can spin one up on rackspace cloud
16:19 ira No idea.
16:19 ira You may just want to use Joyent's SDC cloud.
16:19 johnmark ira: would probably be easier
16:19 johnmark ira: where can I find that?
16:20 ira http://www.joyent.com/ ?
16:20 ira And you'll want a SmartOS instance...
16:21 hagarth johnmark: if we find an illumos box, we can set it up as a jenkins instance.
16:21 ira yeah... there's ways to do this.  I can work with someone to create a KVM VM... etc.
16:21 hagarth johnmark: speaking of which, would it be possible for us to find a public NetBSD instance to run build/smoke tests?
16:22 ira hagarth: Might it be worth setting up a RHEV box somewhere where we can just load all this random stuff up?
16:22 hagarth johnmark: kkeithley_ helped me with a NetBSD VM but unfortunately that lives within RHT and cannot access it from build.gluster.org
16:23 ira Or whatever the open source equivs are...
16:23 johnmark ira: right... I think so
16:23 hagarth ira: yeah, build.gluster.org runs outside RHT atm. So we either need to spin up instances there or have it somewhere where it is reachable.
16:23 * johnmark looks at rackspace cloud info
16:24 ira If you need help bootstraping or understanding Illumos, ping me.  Heaven knows I know it well.
16:25 hagarth ira: sure, I am bound to encounter problems ;)
16:25 ira hagarth: I expect #1 is, what do I run. ;)
16:26 hagarth ira: undoubtedly!
16:28 johnmark heh
16:29 johnmark looks like rackspace cloud only supports freebsd "out of the box"
16:29 johnmark but I have no idea if there's a way to launch your own VM definitions outside of their pre-defined instances
16:31 hagarth johnmark: could jperkin help us with an instance somehow? ;)
16:31 johnmark hagarth: I'm sure he can - joyent has its cloud thing, but I'm not sure how it works
16:31 johnmark I can find out
16:32 kkeithley_ hagarth: hmmm,  I thought you told me once that a Jenkins instance could work from inside RHT? But before we give up on that, let's see if we can open it up to the outside.
16:33 hagarth kkeithley_: making it work from inside RHT would involve running jenkins on that instance.. if it is reachable by the master, then we just need ssh access to the slave.
16:34 hagarth kkeithley_: if the latter is not possible, we would have to explore installing jenkins on NetBSD
16:34 hagarth kkeithley_: looks like NetBSD packages its own jenkins - ftp://ftp.netbsd.org/pub/pkgsrc/curr​ent/pkgsrc/devel/jenkins/README.html
16:36 kkeithley_ I'm not following. What is it we want on the NetBSD VM? Full up Jenkins?
16:36 kkeithley_ Or just something that we can run builds on triggered by the outside Jenkins?
16:37 hagarth kkeithley_: if NetBSD VM acts as a slave to build.g.o, then it doesn't need anything apart from jre.
16:37 kkeithley_ A subset of Jenkins, IOW some kind of Jenkins slave?
16:38 hagarth kkeithley_: if we want the NetBSD VM to act as a separate Jenkins instance, we need full blown installation.
16:38 kkeithley_ okay. Let me investigate exposing it to the outside. We may be able to do that.
16:38 kkeithley_ Whatever works best.
16:38 hagarth kkeithley_: yeah
16:39 kkeithley_ With the storm here lots of people are not in the office, or even on IRC. I may not have an answer until Monday.
16:39 kkeithley_ May not be able to get an answer
16:40 hagarth kkeithley_: heard about the storm, when is it expected to subside?
16:41 kkeithley_ the storm is over. Now everyone is shoveling out
16:42 kkeithley_ Tonight the temp is supposed to go down to -8°F/-22°C
16:44 hagarth wow! I definitely would not survive that. Lots of complaints about the cold around here since temperature has reached 50 F / 12 C.
16:45 johnmark haha
16:46 kkeithley_ It'll be 50°F on Monday — come on over.
16:46 kkeithley_ It's New England. If you don't like the weather, just wait five minutes.
16:46 hagarth kkeithley_: haha
16:53 kkeithley_ but back to the problem at hand. There must be some VPN package(s) we could install on b.g.o and the NetBSD VM, or we could McGyver a reverse ssh tunnel....
16:56 hagarth reverse ssh tunnel might be easier, I'll also try installing jenkins on NetBSD to see if that works too.
17:10 _Bryan_ joined #gluster-dev
20:22 ndk joined #gluster-dev
20:37 johnmark hrm... who did the Debian packages last time?
20:38 johnmark semiosis: ^^
20:38 * semiosis did
20:39 johnmark ah, ok
20:40 semiosis what did i break?
20:41 johnmark semiosis: nothing - just wondering if we can get 3.4.2 builds
20:41 semiosis yes
20:41 johnmark semiosis: ok. I don't like to ask, because I know you'r ebusy
20:41 johnmark but I also don't want to announce until all the builds are there
20:42 semiosis tonight
20:43 johnmark semiosis: thanks :)
20:43 semiosis yw
21:06 kkeithley_ johnmark, semiosis, $whoelse?: I've created a user called glusterpackager on download.gluster.org and created a gpg key with the intention of using this to "officially" sign rpms and dpkgs. N.B. I signed all the rpms I put up today with my personal key, which is what I've been using to date. Does this seem reasonable? Not quite ready to use, I need to upload the key to a keyserver, among other things before it's ready for general u
21:07 semiosis kkeithley_: i made a key i use for the debian packages on d.g.o, deb.packages@gluster.org iirc
21:07 semiosis http://download.gluster.org/pub/gluster​/glusterfs/3.4/3.4.1/Debian/pubkey.gpg
21:08 kkeithley_ oh, okay
21:08 kkeithley_ As long as they're signed by someone I don't suppose it really matters.
21:09 semiosis it's a dedicated key, in the hopes that maybe one day someone else will want to build debs for gluster.org :)
21:09 kkeithley_ ;-)
21:09 kkeithley_ yeah
21:09 kkeithley_ Maybe one day I'll climb that mountain again.
21:10 semiosis now i have a virtualbox vm dedicated to building debs, and a handy readme i wrote for myself which makes it super easy to build them
21:10 semiosis pretty much copying commands from the readme into the VM shell
21:11 semiosis as long as I only have to build for wheezy, that is
21:11 kkeithley_ :-)
21:12 kkeithley_ I forget, is squeeze older that wheezy or is squeeze unstable/bleeding edge? I forget which one I fought with.
21:12 kkeithley_ older than
21:14 kkeithley_ Anyway, if it's super easy, maybe I can take a stab at wrapping it up in a Makefile for an .../extras/LinuxDPKGS/... in the upstream source
21:17 semiosis wheezy is current stable.  squeeze is old/obsolete stable.
21:17 semiosis wheezy will probably be the stable version until mid-2015, assuming ~2 year cycle
21:17 semiosis no one really knows though
21:19 kkeithley_ hmmm. Well, I can't remember. I have a squeeze vm and a wheezy vm. One worked with your (old) recipe, the other one fell down halfway through. ISTR that it was the older dist that fell down.

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