| *** almaisan-away is now known as al-maisan | 05:09 | |
| *** miclorb has joined #openstack | 05:11 | |
| *** davidstrauss has quit IRC | 05:12 | |
| *** _buck has joined #openstack | 05:13 | |
| *** pvo has quit IRC | 05:15 | |
| *** lbieber has joined #openstack | 05:21 | |
| mtaylor | anybody around who can give me a review on https://code.edge.launchpad.net/~mordred/nova/fixes-for-deb-packaging/+merge/31117 ? | 05:31 |
|---|---|---|
| *** BartVB has joined #openstack | 05:35 | |
| *** silassewell has quit IRC | 05:35 | |
| *** BartVB has quit IRC | 05:35 | |
| *** lbieber has quit IRC | 05:44 | |
| *** gnrfan has joined #openstack | 05:51 | |
| vish1 | mtaylor: sounds good | 06:01 |
| vish1 | (i was away at dinner) | 06:01 |
| *** nettok has joined #openstack | 06:04 | |
| *** BartVB has joined #openstack | 06:07 | |
| mtaylor | vish1: awesome. I added the gitignore files back in | 06:18 |
| *** BartVB has quit IRC | 06:23 | |
| *** snowboarder04 has quit IRC | 06:24 | |
| *** gnrfan has quit IRC | 06:25 | |
| *** rajijoom has joined #openstack | 06:26 | |
| mtaylor | vish1: I've got the debian packaging ready to go - I'll get it pushed up in the morning... along with a hudson job to make the packages on trunk pushes | 06:32 |
| vish1 | mtaylor: I guess officially we need one more review before it can go in | 06:33 |
| mtaylor | but for now, I'm going to sleep :) | 06:33 |
| vish1 | mtaylor: sounds good. I've got my hacked version working for tonight. I look forward to doing it properly starting tomorrow | 06:33 |
| *** snowboarder04 has joined #openstack | 06:40 | |
| mtaylor | vish1: excellent | 06:40 |
| *** allsystemsarego has joined #openstack | 06:54 | |
| *** bluenerv1 has joined #openstack | 06:54 | |
| mtaylor | vish1: actually ... | 06:57 |
| mtaylor | vish1: I just realized that I could actually make the packaging branch before the other thing gets merged... | 06:57 |
| mtaylor | vish1: grab lp:~swift-core/nova/debian-packaging and do a bzr bd in it ... and it should Just Work⢠| 06:57 |
| mtaylor | vish1: I made one main packaging change, which was to put the python library in python-nova per-python packaging policy | 06:58 |
| mtaylor | vish1: but I made swift-common depend on that, so it shouldn't seem any different | 06:58 |
| mtaylor | oh - and I added a -doc package for the docs | 06:59 |
| mtaylor | gah | 06:59 |
| mtaylor | I made _nova-common_ depend on it | 06:59 |
| mtaylor | too many projects | 06:59 |
| *** z0 has joined #openstack | 07:07 | |
| *** calavera has joined #openstack | 07:20 | |
| *** nettok has quit IRC | 07:26 | |
| *** metoikos has joined #openstack | 07:32 | |
| *** bluenerv1 has quit IRC | 07:33 | |
| vish1 | mtaylor: bzr bd failed due to missing pgp key. You'll have to let me know how to set that up. But the builddeb.sh script worked | 07:38 |
| mtaylor | vish1: well, that's a good start... | 07:39 |
| mtaylor | vish1: we should add you to uploaders... | 07:39 |
| mtaylor | vish1: do you have a gpg key there? | 07:39 |
| mtaylor | vish1: in any case, I'll write up some notes for you ... | 07:42 |
| *** kashyapc has joined #openstack | 07:43 | |
| *** metoikos has quit IRC | 07:50 | |
| *** metoikos has joined #openstack | 07:53 | |
| *** kashyapc has quit IRC | 08:01 | |
| *** stewart has quit IRC | 08:03 | |
| *** BartVB has joined #openstack | 08:04 | |
| *** miclorb has quit IRC | 08:13 | |
| *** stewart has joined #openstack | 08:14 | |
| *** steph021 has quit IRC | 08:16 | |
| soren | mtaylor: Still here? | 08:23 |
| *** jhalogen has joined #openstack | 08:23 | |
| soren | mtaylor: Good for you :) | 08:23 |
| eday | soren: I have some results of the twisted vs eventlet stuff.. about to post to my blog :) | 08:39 |
| soren | eday: Cool. :) | 08:42 |
| *** hugod has quit IRC | 08:43 | |
| *** hugod has joined #openstack | 08:49 | |
| soren | mtaylor: I've taken the liberty of changing the ownership of the nova debian-packaging branch from swift-core to nova-core. | 08:49 |
| *** fortune has joined #openstack | 09:00 | |
| *** cjwilson has quit IRC | 09:04 | |
| *** adjohn_ has joined #openstack | 09:09 | |
| *** adjohn has quit IRC | 09:10 | |
| *** adjohn_ is now known as adjohn | 09:10 | |
| *** matteo_ is now known as matteo74 | 09:16 | |
| *** z0 has quit IRC | 09:18 | |
| *** yuravk has quit IRC | 09:20 | |
| *** littleidea has quit IRC | 09:25 | |
| *** adjohn_ has joined #openstack | 09:32 | |
| *** adjohn has quit IRC | 09:34 | |
| *** adjohn_ is now known as adjohn | 09:34 | |
| eday | soren: http://oddments.org/?p=494 | 09:36 |
| *** _buck has quit IRC | 09:41 | |
| vish1 | good luck all you europeans! I'm off to bed. See you tomorrow | 09:49 |
| vish1 | fyi all of those patches were needed to actually get the code running properly on our multicomputer setup | 09:49 |
| vish1 | and there may be more tomorrow once i have time to test some more | 09:49 |
| soren | vish1: Whuh, you're still here? | 09:50 |
| soren | vish1: Go to bed! | 09:50 |
| vish1 | soren: i know, right? | 09:50 |
| *** vish1 is now known as vish1-afk | 09:50 | |
| soren | :) | 09:50 |
| *** _buck has joined #openstack | 09:57 | |
| *** _buck has quit IRC | 10:26 | |
| *** matclayton has joined #openstack | 10:27 | |
| *** jonesy_ has quit IRC | 10:30 | |
| *** zebziggle has joined #openstack | 10:30 | |
| *** _buck has joined #openstack | 10:34 | |
| *** zebziggle has left #openstack | 10:39 | |
| *** miclorb_ has joined #openstack | 10:45 | |
| *** jtdowney has joined #openstack | 10:54 | |
| *** z0 has joined #openstack | 10:55 | |
| *** silassewell has joined #openstack | 10:58 | |
| *** _buck has quit IRC | 11:06 | |
| *** jonesy_ has joined #openstack | 11:09 | |
| *** al-maisan is now known as almaisan-away | 11:13 | |
| *** z0 has quit IRC | 11:17 | |
| *** carolinad28 has joined #openstack | 11:42 | |
| *** jtdowney has quit IRC | 11:44 | |
| *** justinsheehy has quit IRC | 11:44 | |
| *** jsmith-away is now known as jsmith | 11:49 | |
| *** miclorb_ has quit IRC | 12:08 | |
| *** miclorb_ has joined #openstack | 12:08 | |
| *** miclorb_ has quit IRC | 12:13 | |
| *** littleidea has joined #openstack | 12:18 | |
| *** justinsheehy has joined #openstack | 12:36 | |
| *** tobym has quit IRC | 12:49 | |
| *** tobym has joined #openstack | 12:50 | |
| *** pvo has joined #openstack | 12:51 | |
| *** pvo has joined #openstack | 12:51 | |
| *** ChanServ sets mode: +v pvo | 12:51 | |
| *** Podilarius has left #openstack | 12:53 | |
| *** tobym has quit IRC | 12:55 | |
| *** tobym has joined #openstack | 13:08 | |
| *** pvo has quit IRC | 13:10 | |
| *** MaxR_ has joined #openstack | 13:24 | |
| *** jbryce has joined #openstack | 13:35 | |
| *** johnpur has joined #openstack | 13:35 | |
| *** ChanServ sets mode: +v johnpur | 13:35 | |
| silassewell | exlt: ping | 13:35 |
| *** sirp2 has quit IRC | 13:38 | |
| *** gein has quit IRC | 13:41 | |
| *** blpiatt has joined #openstack | 13:44 | |
| *** pvo has joined #openstack | 13:49 | |
| *** pvo has joined #openstack | 13:49 | |
| *** ChanServ sets mode: +v pvo | 13:49 | |
| *** abecc has joined #openstack | 13:51 | |
| exlt | silassewell: pong :) | 13:54 |
| *** blpiatt has quit IRC | 13:55 | |
| *** arthurlu has joined #openstack | 13:55 | |
| *** blpiatt has joined #openstack | 13:56 | |
| jonesy | is the swift SAIO still the best doc to do a bunny-hill test installation of swift? I thought there was an installfest doc that showed you how to install it on a cloud server (which I'm using). | 14:01 |
| *** MaxR_ has quit IRC | 14:04 | |
| silassewell | exlt: sorry to keep bugging you, I was wondering what name you were going to submit the debian package as (I'd assume "swift"), just wanted to make sure you guys weren't going to prefix it with something like openstack or anything else (like to keep things somewhat consistent) | 14:05 |
| jsmith | silassewell: FWIW, I thought your spec file looked good | 14:08 |
| jonesy | hm. My cloud server only has a single device. Not gonna work so well for swift testing. | 14:09 |
| *** adjohn has quit IRC | 14:09 | |
| jonesy | :() | 14:09 |
| jonesy | :( | 14:09 |
| *** tobym has quit IRC | 14:09 | |
| exlt | silassewell: that was my thought - 'swift' and 'nova' as the package names (and swift-*, nova-* for the split packages) | 14:10 |
| silassewell | jsmith: thanks, its coming right along, looking forward to testing it on a rack when I get back to work | 14:10 |
| jsmith | exlt: At least in Fedora, there's already a package named "swift" | 14:10 |
| silassewell | exlt: ok, I'll go with that, thanks | 14:10 |
| jsmith | exlt: Might it makes sense to shoot for a package name that doesn't conflict? | 14:11 |
| silassewell | jsmith: it was submitted after my package so I'm pretty safe on the name | 14:11 |
| jsmith | silassewell: Ah, gotcha... | 14:11 |
| *** littleidea has quit IRC | 14:11 | |
| *** littleidea has joined #openstack | 14:12 | |
| exlt | jsmith: there is? debian does not - got a link to the upstream for swift in fedora? | 14:16 |
| exlt | php-channel-swift ? | 14:16 |
| jsmith | exlt: http://swift.im/releases/swift-1.0beta5/ | 14:16 |
| silassewell | exlt: it was a review request submitted after mine https://bugzilla.redhat.com/show_bug.cgi?id=618985 | 14:17 |
| uvirtbot | bugzilla.redhat.com bug 618985 in Package Review "Review Request: Swift - XMPP client" [Medium,Assigned] | 14:17 |
| exlt | gotcha | 14:18 |
| *** almaisan-away is now known as al-maisan | 14:18 | |
| exlt | well, should they go in as openstack-swift[-*]/nova[-*] ? | 14:19 |
| jsmith | exlt: I'd personally prefer that, if that's OK | 14:20 |
| exlt | I'm easy ;-) | 14:20 |
| jsmith | silassewell: You OK w/ that? | 14:21 |
| exlt | jsmith: package names going to distros as openstack-swift and openstack-nova | 14:21 |
| jsmith | silassewell: Or would you rather just use switf/nova without the "openstack-" prefix? | 14:21 |
| *** _mattf has joined #openstack | 14:21 | |
| jsmith | exlt: Yeah | 14:21 |
| silassewell | Fedora will bitch if all they paths don't get updates as well (which could be a pain) | 14:23 |
| *** abecc_ has joined #openstack | 14:23 | |
| silassewell | ex: /etc/init.d/openstack-swift-container vs /etc/init.d/swift-container | 14:23 |
| *** Podilarius has joined #openstack | 14:23 | |
| silassewell | Possibly with the etc path as well, /etc/openstack/swift vs /etc/swift (this seems to be hardcoded all over the place) | 14:24 |
| *** abecc has quit IRC | 14:25 | |
| *** abecc_ is now known as abecc | 14:25 | |
| dendrobates | silassewell: we need to do this for debian/ubuntu as well, but we could do it in packaging, if necessary | 14:26 |
| silassewell | Debian/Ubuntu ends up getting more documentation and tutorials written for it and its nice to be able to use them with fedora (with minimal tweaking), so as long as it stays reasonable consistent I'm fine changing the name | 14:28 |
| *** lbieber has joined #openstack | 14:28 | |
| silassewell | I based the package thus far on the stuff currently done for debian | 14:29 |
| silassewell | Looks like FreeBSD already has a nova package although swift is free, maybe its better just to prefix it | 14:31 |
| *** mray has joined #openstack | 14:32 | |
| *** ianw has joined #openstack | 14:32 | |
| exlt | silassewell: sounds like it - I hadn't checked the *BSDs | 14:32 |
| *** sophiap has joined #openstack | 14:36 | |
| *** calavera has quit IRC | 14:40 | |
| *** lbieber has quit IRC | 14:41 | |
| notmyname | jonesy: the SAIO doc is best for testing (very) small usage. the only thing to change for a cloud server is to use a loopback device | 14:43 |
| *** tr3buchet has joined #openstack | 14:44 | |
| *** tr3buchet has joined #openstack | 14:45 | |
| mtaylor | soren: thanks :) | 14:49 |
| mtaylor | soren: I was starting to have swift/nova naming issues last night | 14:50 |
| soren | mtaylor: I've fixed up the packaging some. | 14:50 |
| mtaylor | soren: excellent | 14:50 |
| soren | mtaylor: ...and I've pushed a few of the missing dependencies to Maverick. | 14:50 |
| mtaylor | soren: double excellent | 14:50 |
| soren | mtaylor: We're on good track to get in before FF. | 14:50 |
| soren | mtaylor: I expect to upload the first version to universe tomorrow. I just want to see it build in a PPA first, then I'm happy. | 14:51 |
| *** pvo_ has joined #openstack | 14:52 | |
| *** sophiap has quit IRC | 14:52 | |
| creiht | eday: I'm curious to hear about your results | 14:57 |
| creiht | jonesy: there is a link to a chef script on the openstack wiki that may auto-install the things you need | 14:57 |
| *** jbryce has quit IRC | 14:57 | |
| jonesy | oh nice. | 14:57 |
| jonesy | thanks | 14:57 |
| soren | creiht: 09:36 < eday> soren: http://oddments.org/?p=494 | 14:58 |
| creiht | jonesy: btw: on a cloud server, you can set an extra partition as a loopback device | 14:58 |
| soren | (assuming you mean those results) | 14:58 |
| *** jbryce has joined #openstack | 14:58 | |
| creiht | soren: thanks | 14:59 |
| *** adrian_otto has joined #openstack | 15:00 | |
| jonesy | creiht: sure thing - thanks. | 15:01 |
| soren | mtaylor: Debian is shaping up reasonably well, too. I've sent python-gflags to my sponsor, I've filed a bug report against python-redis to get it updated to 2.0.0.. | 15:02 |
| soren | mtaylor: It's just easier for me to work with Ubuntu, of course. | 15:02 |
| creiht | eday: btw, with your argument about utilizing cores, for network services like we are doing, it is very easy to scale to cores if you use a shared nothing architecture and the services fork to however many workers you want | 15:03 |
| creiht | We do this with great effect in swift | 15:03 |
| mtaylor | soren: yes. of course. perhaps I'll actually become a DD next week at debconf (not really holding my breath, but you never know) | 15:03 |
| soren | mtaylor: Sweet. Then I'll leave nova itself to you. :) | 15:04 |
| *** blpiatt has quit IRC | 15:04 | |
| * mtaylor needs more upload rights to more things... | 15:06 | |
| *** sirp1 has joined #openstack | 15:06 | |
| *** MaxR_ has joined #openstack | 15:06 | |
| *** blpiatt has joined #openstack | 15:06 | |
| *** adrian_otto has quit IRC | 15:10 | |
| *** pcrews has joined #openstack | 15:11 | |
| mtaylor | soren: piddle. ${python:Depends} for some reason did not _actually_ fill in the depends | 15:12 |
| soren | mtaylor: Which depends? | 15:12 |
| mtaylor | soren: for python-nova | 15:12 |
| soren | mtaylor: Which depends? :) | 15:12 |
| soren | mtaylor: The ones from build-dep? | 15:12 |
| mtaylor | soren: just normal installation depends | 15:12 |
| *** tobym has joined #openstack | 15:13 | |
| soren | mtaylor: Still not following :) | 15:13 |
| mtaylor | soren: debian/control line 23 | 15:13 |
| mtaylor | soren: I removed things that were explicit there, thinking that python:Depends would be a good boy and fill them in for us ... but it did not | 15:14 |
| soren | mtaylor: Yes, sure, but which dependencies where you expecting to turn up thre? | 15:14 |
| soren | mtaylor: python:Depends isn't magical live shlibs depends. | 15:14 |
| mtaylor | soren: ah. that's the problem then | 15:14 |
| creiht | mtaylor: btw: If the swift packages in the ppa are not working, we probably shouldn't have them up there | 15:15 |
| mtaylor | creiht: yes. I agree. will fix this morning | 15:15 |
| creiht | cool thanks | 15:15 |
| soren | mtaylor: At the moment, python:Depends only ever really adds dependencies for specific version of python itself. | 15:17 |
| soren | mtaylor: Like if a script specifically says #!/usr/bin/python2.4 at the top. | 15:18 |
| *** jbryce has quit IRC | 15:18 | |
| mtaylor | soren: yeah. now that you say that my brain got better... for some reason it was, well, dumb last night | 15:18 |
| soren | mtaylor: I did push the stack of b-d's, right? | 15:18 |
| *** jbryce has joined #openstack | 15:19 | |
| * soren hugs "bzr lp-open" | 15:19 | |
| mtaylor | soren: ooh! I haven't seen that command yet... | 15:19 |
| soren | It opens the lp page for the branch corresponding to your $CWD | 15:19 |
| *** dendrobates is now known as dendro-afk | 15:19 | |
| mtaylor | soren: that's so awesome | 15:19 |
| soren | "bzr lp-open lp:nova" also works, for instance. | 15:19 |
| soren | Very awesome indeed. | 15:20 |
| soren | An "very awesome" is pretty fscking awesome. | 15:20 |
| mtaylor | ++ | 15:20 |
| *** dendro-afk is now known as dendrobates | 15:22 | |
| * soren heads out | 15:26 | |
| *** fortune has quit IRC | 15:29 | |
| mtaylor | ok. I think we're set for nova packaging - will fix swift packaging next (and the automation around both) | 15:33 |
| mtaylor | but first - I must bathe... | 15:33 |
| *** sophiap has joined #openstack | 15:35 | |
| *** ianweller has joined #openstack | 15:35 | |
| *** sophiap has quit IRC | 15:39 | |
| rbergeron | ianweller ;) | 15:45 |
| ianweller | o hai | 15:45 |
| *** bluenerv has joined #openstack | 15:52 | |
| *** pvo_ has quit IRC | 15:55 | |
| *** bluenerv has quit IRC | 15:56 | |
| *** bluenerv has joined #openstack | 15:58 | |
| *** mtaylor has quit IRC | 16:01 | |
| jsmith | ianweller: You're here too? I can't escape from you! | 16:04 |
| *** metoikos has quit IRC | 16:10 | |
| *** metoikos has joined #openstack | 16:15 | |
| *** bluenerv has quit IRC | 16:22 | |
| *** jdmaturen has joined #openstack | 16:25 | |
| *** jzb has joined #openstack | 16:28 | |
| *** jdmaturen has quit IRC | 16:30 | |
| *** jdmaturen has joined #openstack | 16:30 | |
| *** bluenerv has joined #openstack | 16:32 | |
| *** bluenerv has joined #openstack | 16:39 | |
| *** mtaylor has joined #openstack | 16:41 | |
| *** ChanServ sets mode: +v mtaylor | 16:41 | |
| mtaylor | back | 16:42 |
| *** lbieber has joined #openstack | 16:46 | |
| mtaylor | creiht, lbieber: any idea if joelbm24 has signed the CLA? | 16:49 |
| lbieber | mtaylor: - http://wiki.openstack.org/Approved%20Contributors | 16:50 |
| mtaylor | lbieber: ok. nobody named joel on there | 16:51 |
| dendrobates | my oscon talk is up for those who missed it. http://goo.gl/n2KJ | 16:51 |
| mtaylor | lbieber: I _really_ need to get the launchpad integration thing going | 16:51 |
| *** arthurlu has quit IRC | 16:55 | |
| *** sirp1 has quit IRC | 17:03 | |
| *** blpiatt has left #openstack | 17:08 | |
| *** blpiatt has joined #openstack | 17:08 | |
| *** StylusEater_work has joined #openstack | 17:10 | |
| StylusEater_work | hi. I'm trying to fix bug 610583 but during testing of my patch I realized I can't invoke swift-account-audit because of an import failure. I tracked it down to from swift.common.ring import Ring. The package file in the ring directory seemed weird... it contained import statements ... is this an issue or am I not understanding the intention of the __init__.py under the ring directory? | 17:13 |
| uvirtbot | Launchpad bug 610583 in swift "standardize on optparse" [Low,New] https://launchpad.net/bugs/610583 | 17:13 |
| *** ianw has quit IRC | 17:14 | |
| eday | creiht: Yeah, that's one of the points I was trying to make near the bottom.. run multiple instances per server or push to worker process pools. Not all services are a good fit for this model | 17:14 |
| *** algorist has joined #openstack | 17:15 | |
| *** mcgurrin has joined #openstack | 17:16 | |
| mcgurrin | is the split/join fest over? | 17:16 |
| mcgurrin | it was stopping me from working with the notifications it was sending | 17:17 |
| *** mray has quit IRC | 17:21 | |
| eday | creiht: did you guys look at gevent? http://www.gevent.org/ someone responded that it is cleaner than eventlet | 17:22 |
| *** _buck has joined #openstack | 17:26 | |
| *** davidstrauss has joined #openstack | 17:28 | |
| StylusEater_work | found my answer | 17:29 |
| *** _mattf has quit IRC | 17:30 | |
| chromakode | mtaylor, derp! what did I miss? | 17:48 |
| chromakode | I think you pinged me exactly when I got netsplit | 17:48 |
| mtaylor | chromakode: perms on irc logs... | 17:48 |
| chromakode | oh gah :) | 17:48 |
| *** al-maisan is now known as almaisan-away | 17:50 | |
| chromakode | I think there must be some code in bip that's setting the permissions, because the directory has the sticky bit set | 17:52 |
| *** _buck has quit IRC | 17:52 | |
| chromakode | we could: 1. be silly and cron it, 2. be smart and patch it, 3. use a different logger | 17:52 |
| mtaylor | chromakode: mmm | 17:53 |
| mtaylor | chromakode: soo.... | 17:53 |
| *** vish1-afk is now known as vish1 | 17:54 | |
| vish1 | who knows how to scroll up in irssi? | 17:54 |
| *** _buck has joined #openstack | 17:55 | |
| *** _mattf has joined #openstack | 17:56 | |
| chromakode | man irssi does ;) | 18:02 |
| * mtaylor pokes chromakode in the eye | 18:02 | |
| eday | vish1: page-up | 18:03 |
| eday | vish1: what were your main reasons to switch away from tornado? | 18:06 |
| *** matclayton has left #openstack | 18:07 | |
| comstud | they cause too much destruction | 18:07 |
| *** blpiatt has quit IRC | 18:12 | |
| lbieber | mtaylor: still no news on the swift mailing list?? | 18:19 |
| mtaylor | lbieber: they're still looking in to it | 18:20 |
| mtaylor | lbieber: sorry I don't have better news on that front | 18:20 |
| lbieber | mtaylor: it just seems that it can't be that difficult to fix!! Pretty frustrating | 18:20 |
| mtaylor | lbieber: they are also looking in to the weird perms on launchpad.net/swift - but I think they've found the issue with that one | 18:20 |
| *** sirp2 has joined #openstack | 18:26 | |
| *** rajijoom has quit IRC | 18:34 | |
| vish1 | eday: anyone know how to send page-up from mac over ssh? | 18:38 |
| gholt | eday: We looked at gevent; it started up after we were already in-progress. Theoretically it should be an easy port if desired. | 18:38 |
| vish1 | eday: minimize dependencies | 18:38 |
| *** mtaylor has quit IRC | 18:38 | |
| vish1 | eday: we were using all of the deferred stuff from twisted and patching ioloop to work with it | 18:39 |
| gholt | vish1: Fn-Up doesn't do it for you? | 18:39 |
| vish1 | eday: and the specific combination of evented systems, forking, ioloop was a bit nasty | 18:39 |
| gholt | [btw, you should definitely say fn-up out loud if possible] | 18:39 |
| vish1 | eday: and since we have very little need for web services we're cutting out tornado | 18:40 |
| vish1 | eday: also tornado couldn't serve multigig files | 18:40 |
| creiht | mtalor isn't here, but I thought Joel had signed the CLA :/ | 18:41 |
| vish1 | gholt: thanks now i'm definitely func-tion-up things | 18:41 |
| *** lbieber has quit IRC | 18:41 | |
| creiht | StylusEater_work: Give me just a sec, and I will look into the import issue | 18:42 |
| creiht | eday: re: gevent, long story, but it was a fork of eventlet, and of course you are going to have people on both side, of which I fall on the eventlet side | 18:44 |
| *** jsmith has quit IRC | 18:44 | |
| *** jsmith has joined #openstack | 18:44 | |
| StylusEater_work | +creiht: not certain if it's a problem with the glibc on the mac | 18:45 |
| StylusEater_work | +creiht: b/c I'm getting a AttributeError: dlsym(0x8fe31d18, fallocate): symbol not found error | 18:45 |
| notmyname | isn't fallocate a linux thing? | 18:46 |
| eday | vish1: ahh, ok, thanks | 18:46 |
| creiht | StylusEater_work: ahh yeah, we may have some things in there that are linux specific | 18:46 |
| eday | gholt: ok. on my blog comments someone suggested gevent had better performance than eventlet, was just curious if you guys had seen otherwise | 18:47 |
| eday | creiht: yeah, makes sense :) | 18:47 |
| StylusEater_work | +creiht: trying to repeat on my debian lenny machine... but now I'm running into a different issue... python versioning ... haha | 18:47 |
| creiht | eday: been a while since I have looked at gevent, but it is built on top of libevent | 18:48 |
| creiht | Which it gets most of its speed from by using libevent's http server | 18:48 |
| creiht | It also at the time had its own quirks, like you couldn't use threads with it | 18:49 |
| creiht | It is also funny, that (iirc), most of the issues that he had with eventlet, have since been fixed | 18:51 |
| creiht | but as most things go, He of course knows how to do everything better :) | 18:52 |
| * creiht also notes that he will be a bit opinionated on the topic | 18:52 | |
| * gholt wonders why creiht capitalized "He" that one time... | 18:53 | |
| creiht | Not sure if gevent is used in any large hi-throughput projects, but eventlet is used at Second Life, and in CloudFiles | 18:53 |
| creiht | anyways... I'll get off my soap box :) | 18:54 |
| cory_ | gholt: nice | 18:54 |
| StylusEater_work | +creiht: what distro do you develop on? | 18:54 |
| creiht | StylusEater_work: Ubuntu 10.04 | 18:54 |
| StylusEater_work | +creiht: 'fraid of that...k thanks | 18:54 |
| creiht | StylusEater_work: The main requirement is python 2.6 at the moment | 18:55 |
| gholt | Even the linux-specific stuff I don't think *needs* to be in there; we just don't have nice checking to skip over the linuxy optimizations. | 18:56 |
| eday | creiht: hehe | 18:56 |
| StylusEater_work | +creiht: picked up on that... thanks | 18:56 |
| * creiht goes to add a bug to make swift 2.5 compatible | 18:57 | |
| dubs | /./. | 18:58 |
| dubs | oops | 18:58 |
| StylusEater_work | +creiht: probably not a biggie since 3000 is now the new "standard" | 18:58 |
| StylusEater_work | err 3.0 (aka 3000) | 18:59 |
| creiht | StylusEater_work: Well a lot of distros still have 2.5 as the standard python version | 18:59 |
| creiht | and it should be a fairly easy fix | 18:59 |
| StylusEater_work | +creiht: good point. | 19:00 |
| StylusEater_work | +creiht: on debian 5.0.5 ... /lib/i686/cmov/libc.so.6: undefined symbol: fallocate | 19:02 |
| creiht | interesting | 19:02 |
| creiht | redbo: -^ | 19:03 |
| creiht | StylusEater_work: digging... | 19:03 |
| *** tonywolf has joined #openstack | 19:04 | |
| StylusEater_work | +creiht: bet you has to do with the libc version... again | 19:04 |
| uvirtbot | New bug: #611007 in swift "Make Swift Python 2.5 compatible" [Undecided,New] https://launchpad.net/bugs/611007 | 19:05 |
| StylusEater_work | +creiht: updating libc to unstable to see if it's an option enabled in a newer version... which I imagine ubuntu would already have | 19:09 |
| creiht | StylusEater_work: Still seems weird to me, since I thought fallocate had been available for a while | 19:09 |
| * creiht is still digging | 19:09 | |
| creiht | StylusEater_work: When I do a man fallocate, I see the following at the very bottom: http://pastebin.com/HTAur1F5 | 19:11 |
| creiht | perhaps that has something to do with it? | 19:11 |
| creiht | well that is for the command line | 19:13 |
| creiht | heh | 19:13 |
| StylusEater_work | +creiht: oh no... I didn't RTFM ... *gasp* :-) | 19:13 |
| StylusEater_work | +creiht: testing | 19:14 |
| creiht | provided in glibc 2.10 | 19:14 |
| creiht | (from man 2 fallocate) | 19:15 |
| StylusEater_work | +creiht: debian 5.0.5 I believe is at glibc-2.7-1 | 19:19 |
| creiht | hrm | 19:20 |
| chmouel | posix_fallocate was introduced around kernel 2.6.23 | 19:20 |
| chmouel | at least the kernel side | 19:20 |
| chmouel | sorry fallocate in kernel is faster than the posix_fallocate from glibc | 19:21 |
| creiht | StylusEater_work: So the other option, is that we could add a config, (or at least detect if it isn't available), and not fallocate | 19:22 |
| *** tobym has quit IRC | 19:22 | |
| creiht | though not fallocating means that you are likely to see files on several extents | 19:23 |
| * creiht adds a bug | 19:25 | |
| * StylusEater_work hopes he can help close bugs he's finding | 19:30 | |
| creiht | :) | 19:31 |
| jonesy | creiht: got a link to that chef script? Searched the wiki and sifted through some links there and can't find it. | 19:31 |
| creiht | jonesy: http://wiki.openstack.org/GettingTheCode | 19:32 |
| creiht | under Swift All In One (Auto Install) | 19:33 |
| StylusEater_work | +creiht: neat... using chef to autoprovision a dev box? | 19:33 |
| creiht | StylusEater_work: yes | 19:33 |
| jonesy | creiht: the one link I didn't click :-/ | 19:33 |
| jonesy | thank you | 19:33 |
| creiht | holoway: When are you going to send a merge request to add your chef script to the source repo? :) | 19:34 |
| jonesy | creiht: also, might wanna mention the word 'chef' in there somewhere so people using the search box might find it more easily. | 19:34 |
| creiht | Good point | 19:34 |
| *** ianw has joined #openstack | 19:36 | |
| *** mrayzenoss has joined #openstack | 19:36 | |
| *** mrayzenoss has joined #openstack | 19:36 | |
| creiht | done | 19:37 |
| jonesy | nice. | 19:38 |
| jonesy | my install is running :) | 19:38 |
| creiht | woot | 19:38 |
| StylusEater_work | +creiht: ok just built and installed util-linux-ng ... works | 19:39 |
| creiht | ahh... interesting | 19:39 |
| StylusEater_work | +creiht: so there is a "workaround" ... if you want to call it that | 19:39 |
| creiht | StylusEater_work: in building util-linux-ng, did it also replace glibc? | 19:40 |
| *** mtaylor has joined #openstack | 19:40 | |
| *** ChanServ sets mode: +v mtaylor | 19:40 | |
| uvirtbot | New bug: #611019 in swift "Gracefully handle unavailability of fallocate system call" [Undecided,New] https://launchpad.net/bugs/611019 | 19:41 |
| *** carolinad28 has quit IRC | 19:41 | |
| creiht | mtaylor: I'm not sure who joelbm24 is? | 19:41 |
| StylusEater_work | +creiht: dpkg -l | grep libc6 ... 2.11.2-2 ... so yes | 19:41 |
| creiht | hah | 19:41 |
| mtaylor | creiht: me either - he had a merge prop in ... I bounced it and asked him to sign the CLA | 19:41 |
| creiht | then you could have just rebuilt glibc :p | 19:42 |
| creiht | mtaylor: swift? | 19:42 |
| StylusEater_work | +creiht: although it's funny dpkg lists it as updated... guess it just reads the symlink then head of the lib | 19:43 |
| mtaylor | creiht: uh, no, actually it was in nova | 19:43 |
| creiht | mtaylor: then why are you asking me? :) | 19:43 |
| StylusEater_work | +creiht: b/c the symlink for libc.so.6 now points to the 2.11.2-2 lib under /lib | 19:44 |
| creiht | StylusEater_work: hah | 19:44 |
| creiht | That probably isn't a solution that your average debian user is going to like :) | 19:44 |
| *** tobym has joined #openstack | 19:44 | |
| StylusEater_work | +creiht: but rebuilding glibc... :-/ ... guess it's just as bad as install util-linux-ng | 19:44 |
| mtaylor | creiht: because you know _everything_ ! | 19:45 |
| StylusEater_work | +creiht: yes. the average *nix user wouldn't like it much but I wouldn't say openstack is for the average user | 19:45 |
| StylusEater_work | +mtaylor: ha | 19:46 |
| jonesy | haha - I have to hack my ubuntu vm. Tried to add myself to a new group using usermod -g instead of -G, so now that's my only group, and apparently sudoers uses group membership, not username. | 19:46 |
| StylusEater_work | +mtaylor: +creiht doesn't? :-p | 19:46 |
| *** ianw has quit IRC | 19:46 | |
| mtaylor | jonesy: yes. that - um .... fail | 19:46 |
| StylusEater_work | yikes | 19:47 |
| StylusEater_work | sorry jonesy | 19:47 |
| jonesy | mtaylor: totally - but at least it's not like, total fail. | 19:47 |
| mtaylor | indeed | 19:47 |
| StylusEater_work | jonesy: runlevel 1 it? | 19:47 |
| mtaylor | it could have caused evil leprechauns to flood out of your cd drive and kill you | 19:48 |
| StylusEater_work | +creiht: I don't have a swift env setup... mind testing something? | 19:48 |
| mtaylor | WHY DOES SOLARIS SUCK? | 19:48 |
| mtaylor | sorry. just had to get that off my chest there | 19:49 |
| StylusEater_work | +creiht: diff... http://pastebin.com/NVs5iFE3 | 19:49 |
| StylusEater_work | +mtaylor: umm it doesn't | 19:49 |
| mtaylor | StylusEater_work: I promise, this isn't an argument we want to get in to at this precise moment... although I allow that you might have a valid point | 19:50 |
| uvirtbot | New bug: #611022 in swift "background of KDE pannel is half-black when using 100% of screen width" [Undecided,New] https://launchpad.net/bugs/611022 | 19:51 |
| StylusEater_work | +mtaylor: just figured I'd give a dissenting opinion ... I've used it... it has nice features ... but yes... don't need to get into it | 19:51 |
| mtaylor | hehe | 19:51 |
| mtaylor | well, normally I'd find it fun to debate, but currently it's giving me build nightmares | 19:51 |
| StylusEater_work | +mtaylor: ahh yes... setting up buildbot on sol10 for mariadb wasn't fun for me ... ran into a whole host of issues | 19:52 |
| mtaylor | StylusEater_work: hrm... mariadb... feels like it's likely I know you then? | 19:53 |
| StylusEater_work | +mtaylor: maybe ... I run three buildbots for mariadb | 19:53 |
| StylusEater_work | +mtaylor: beyond that ... no commits quite yet | 19:54 |
| mtaylor | StylusEater_work: ah, but you aren't a monty program/former mysql person | 19:54 |
| mtaylor | StylusEater_work: cool.... well thanks for contributing buildbot slaves to them! (and especially for contributing solaris... *shudder*) | 19:54 |
| StylusEater_work | +mtaylor: haven't had the privilege | 19:55 |
| *** brainproxy has quit IRC | 19:59 | |
| *** jdmaturen has quit IRC | 20:01 | |
| *** jdmaturen has joined #openstack | 20:03 | |
| *** tobym_ has joined #openstack | 20:06 | |
| *** tobym has quit IRC | 20:06 | |
| creiht | StylusEater_work: Are you running into other issues other than the fallocate thing? | 20:09 |
| *** kallistec has quit IRC | 20:10 | |
| StylusEater_work | +creiht: another diff ... http://pastebin.com/Dd90LRnX | 20:11 |
| StylusEater_work | +creiht: no | 20:11 |
| StylusEater_work | +creiht: but I'm not really "running it yet" ... just fixing the getopt bug atm. | 20:11 |
| *** brainproxy has joined #openstack | 20:12 | |
| creiht | StylusEater_work: What would be easier, would be to push your branch as a local repo | 20:12 |
| creiht | gotcha | 20:12 |
| creiht | StylusEater_work: someting like bzr push lp:~user_id/swift/lp610583 | 20:12 |
| creiht | Then I can just do a checkout | 20:12 |
| * StylusEater_work bit of a bzr newb | 20:14 | |
| creiht | http://wiki.openstack.org/LifeWithBzrAndLaunchpad | 20:14 |
| *** rnewson has joined #openstack | 20:15 | |
| creiht | and you should also take a look at http://wiki.openstack.org/HowToContribute | 20:15 |
| creiht | so you can sign the CLA if you would like to contribute... it is quite painless (I promise) | 20:15 |
| *** jbryce has quit IRC | 20:17 | |
| *** jbryce has joined #openstack | 20:17 | |
| StylusEater_work | +creiht: already signed and completed that stuff on announcement | 20:18 |
| creiht | mtaylor: what is proper etiquette for when someone posts a bug for a wrong project? | 20:19 |
| creiht | https://bugs.launchpad.net/swift/+bug/611022 | 20:19 |
| uvirtbot | Launchpad bug 611022 in swift "background of KDE pannel is half-black when using 100% of screen width" [Undecided,New] | 20:19 |
| creiht | StylusEater_work: awesome | 20:19 |
| mtaylor | creiht: you can just reassign it to a different project | 20:19 |
| mtaylor | creiht: I'm gonna guess something kde releated :) | 20:19 |
| creiht | StylusEater_work: Would it be possible to follow the launchpad guide and make a merge proposal for your changes? | 20:20 |
| mtaylor | creiht: done | 20:22 |
| creiht | mtaylor: thanks! | 20:23 |
| StylusEater_work | +creiht: yes | 20:24 |
| StylusEater_work | +creiht: was reading through that... not sure how quickly I can do it | 20:24 |
| creiht | StylusEater_work: cool | 20:24 |
| creiht | StylusEater_work: No rush, but it makes it easier for all in the end | 20:24 |
| creiht | basic idea is to branch swift locally, make your change, commit, push branch to your user's repo, and request a merge | 20:25 |
| mtaylor | gonna be so much nicer when the lp-submit command is in mainline so we can start including that in standard instructions... | 20:25 |
| * creiht is looking forward to that | 20:26 | |
| *** lbieber has joined #openstack | 20:26 | |
| StylusEater_work | +creiht: so far I did ... bzr branch lp:swift ... made changes ... | 20:27 |
| creiht | StylusEater_work: then you need to commit the change locally | 20:27 |
| StylusEater_work | +creiht: I didn't bzr init ... bzr branch ... | 20:27 |
| creiht | StylusEater_work: that's fine | 20:27 |
| creiht | bzr init just makes it so that bzr info is shared among branches | 20:27 |
| mtaylor | s/bzr init/bzr init-repo/ ... bzr init actually makes a new empty branch | 20:28 |
| creiht | oh yeah :) | 20:29 |
| * creiht is still a bit new to bzr as well | 20:29 | |
| *** justinsheehy has quit IRC | 20:32 | |
| *** miclorb has joined #openstack | 20:34 | |
| *** lbieber has quit IRC | 20:38 | |
| *** _mattf has quit IRC | 20:39 | |
| StylusEater_work | +creiht: are you PST? | 20:39 |
| creiht | StylusEater_work: Central | 20:40 |
| *** rnewson has quit IRC | 20:40 | |
| *** rnewson has joined #openstack | 20:42 | |
| *** lbieber has joined #openstack | 20:45 | |
| *** jtdowney has joined #openstack | 20:45 | |
| *** jonesy_ has quit IRC | 20:49 | |
| StylusEater_work | +creiht: bbiab | 20:49 |
| *** gundlach has joined #openstack | 20:50 | |
| *** StylusEater_work has left #openstack | 20:50 | |
| *** kallistec has joined #openstack | 20:52 | |
| *** MaxR_ has quit IRC | 20:54 | |
| *** cf_conrad has joined #openstack | 21:00 | |
| *** cf_conrad is now known as cweidenkeller | 21:01 | |
| *** justinsheehy has joined #openstack | 21:04 | |
| *** JimCurry has joined #openstack | 21:12 | |
| *** JimCurry has left #openstack | 21:14 | |
| *** JimCurry has joined #openstack | 21:14 | |
| *** JimCurry has left #openstack | 21:15 | |
| *** sander__ has quit IRC | 21:15 | |
| *** JimCurry has joined #openstack | 21:15 | |
| *** justinsheehy has quit IRC | 21:18 | |
| *** sander__ has joined #openstack | 21:19 | |
| mtaylor | creiht, redbo: k. new debs pushed to ppa ... lets see if these are any better | 21:25 |
| creiht | mtaylor: cool... I'm having letterj check it out | 21:26 |
| *** justinsheehy has joined #openstack | 21:27 | |
| *** allsystemsarego has quit IRC | 21:28 | |
| *** _buck has quit IRC | 21:31 | |
| creiht | mtaylor: If I wanted to build the package myself, how would that be done? | 21:34 |
| mtaylor | creiht: bzr branch lp:~swift-core/swift/debian ; cd debian ; bzr bd | 21:36 |
| *** miclorb has quit IRC | 21:36 | |
| mtaylor | creiht: it make take a bit before launchpad has built the debs and published them to the ppa | 21:37 |
| creiht | k | 21:37 |
| *** miclorb has joined #openstack | 21:37 | |
| mtaylor | creiht: also - you might need to apt-get install bzr-builddeb :) | 21:37 |
| *** jsmith is now known as jsmith-away | 21:38 | |
| *** gundlach has quit IRC | 21:40 | |
| *** miclorb has quit IRC | 21:41 | |
| letterj | mtaylor: I installed bzr-builddeb and ran bzr bd. It errored out | 21:41 |
| *** BartVB has quit IRC | 21:42 | |
| letterj | mtaylor: bzr: ERROR: Unable to find the needed upstream tarball: swift_20100728203750.orig.tar.gz. | 21:42 |
| mtaylor | letterj: hrm. that's disturbing | 21:42 |
| mtaylor | letterj: lemme try from scratch somewhere ... | 21:43 |
| mtaylor | letterj: hrm. worked for me - which is even more troubling (would have prefered it break for me) | 21:45 |
| mtaylor | letterj: what version do you have installed? (bzr plugins | grep builddeb) | 21:46 |
| letterj | there is an option: --export-upstream=ARG Create the .orig.tar.gz from a bzr branch before building. | 21:47 |
| letterj | builddeb 2.2.0 | 21:47 |
| letterj | what options did you add to brz db | 21:48 |
| letterj | sorry /db/bd/ | 21:50 |
| *** miclorb_ has joined #openstack | 21:50 | |
| gholt | For me, it just has problems at the signing stage since I'm not mordred@inaugust.com I guess. :) | 21:51 |
| mtaylor | letterj: none. just bzr bd | 21:53 |
| mtaylor | gholt: you can get around that with "bzr bd --bulder='debuild -uc -us' | 21:53 |
| gholt | Ah, thanks. | 21:53 |
| mtaylor | letterj: for you ... it _should_ be creating the tarball all automagically for you | 21:53 |
| gholt | They should put that in the bzr bd --help | 21:53 |
| mtaylor | letterj: do you have pristine-tar installed? | 21:54 |
| gholt | Cool, worked for me. | 21:54 |
| gholt | Note: I haven't actually /used/ the packages. | 21:55 |
| mtaylor | gholt: heh. python-swift _seems_ to have actually installed the python files this time | 21:56 |
| letterj | yep | 21:57 |
| mtaylor | hrm | 21:59 |
| mtaylor | letterj: I'm pulling at straws here ... what v of bzr are you running? | 22:01 |
| * mtaylor is running 2.1.1 | 22:01 | |
| * mtaylor has also asked over in #bzr | 22:02 | |
| letterj | started back from scratch and repulled the code. A dependency error came up for python-all. Fixed that and everything worked ok | 22:03 |
| mtaylor | letterj: heh. how weird | 22:04 |
| mtaylor | well good! | 22:04 |
| *** johnpur has quit IRC | 22:05 | |
| *** rnewson has quit IRC | 22:05 | |
| *** StylusEater has joined #openstack | 22:05 | |
| StylusEater | +creiht: still around? | 22:05 |
| creiht | StylusEater: indeed | 22:07 |
| letterj | The --builder='debuild -uc -us' is was the thing I needed. Once I had that the real error was visible | 22:07 |
| *** jonesy_ has joined #openstack | 22:10 | |
| mtaylor | letterj: ah, excellent | 22:10 |
| *** jonesy_ has quit IRC | 22:11 | |
| mtaylor | letterj: you can put that in ~/.bazaar/builddeb.conf | 22:11 |
| mtaylor | letterj: mine looks like: http://pastebin.com/qzJDhRHa | 22:11 |
| *** tobym_ has quit IRC | 22:14 | |
| letterj | mtaylor: I wanted to just document these steps somewhere. Where should I do that? | 22:15 |
| mtaylor | letterj: well, I've got a wiki page up for maintaining the package... | 22:16 |
| mtaylor | letterj: perhaps on that page | 22:16 |
| mtaylor | letterj: http://wiki.openstack.org/DebianPackaging | 22:16 |
| *** davidstrauss has quit IRC | 22:25 | |
| *** tobym has joined #openstack | 22:35 | |
| *** rnewson has joined #openstack | 22:36 | |
| *** jonesy_ has joined #openstack | 22:38 | |
| StylusEater | +creiht: should I create a wiki page on howto setup a debian lenny dev box? | 22:41 |
| StylusEater | +creiht: or is that done already? | 22:41 |
| mtaylor | can I get somebody to give me an ok on https://code.edge.launchpad.net/~mordred/nova/generate-changelog/+merge/31211 | 22:49 |
| mtaylor | and also on https://code.edge.launchpad.net/~mordred/swift/generate-changelog/+merge/31209 | 22:50 |
| *** sander__ has quit IRC | 22:51 | |
| *** sander__ has joined #openstack | 22:52 | |
| redbo | I should totally reject that for pep8 | 22:57 |
| *** JimCurry has left #openstack | 22:58 | |
| *** lbieber has quit IRC | 22:58 | |
| mtaylor | damn | 23:00 |
| mtaylor | redbo: ok. fixed pep8 in setup.py... :) | 23:02 |
| vish1 | mtaylor: i ok'd it but i have no idea what kind of voodoo it is doing | 23:03 |
| mtaylor | vish1: :) ... it's just making a ChangeLog file from bzr log before we make a source tarball | 23:03 |
| mtaylor | unlikely that anyone will ever notice it happens :) | 23:03 |
| vish1 | mtaylor: that is great though...editing changelog by hand is annoying | 23:05 |
| *** pvo has quit IRC | 23:05 | |
| redbo | So the merge request doesn't seem to reflect the new change to the branch. When I approve it does the whole branch go in, or only what was proposed? | 23:05 |
| mtaylor | vish1: yes. especially when you also have commit message already | 23:05 |
| vish1 | mtaylor: so if i'm building a local branch how exactly do i keep it from conflicting with the public ppa | 23:06 |
| mtaylor | redbo: it will only be merge up to the revision the branch was on at the time its status was set to Approved | 23:06 |
| redbo | oh I see, it says it's udpating the diff now | 23:06 |
| mtaylor | vish1: uh... you mean if you build/install packages locally and you also have the ppa in your sources.list ? | 23:07 |
| mtaylor | vish1: you could bump your local revision in debian/changelog so that it's higher than what's going in the ppa | 23:07 |
| vish1 | mtaylor: i have to manage an internal deploy branch | 23:08 |
| mtaylor | ah. gotcha | 23:08 |
| redbo | I always put my review type as "awesome". Is that right? | 23:08 |
| mtaylor | vish1: then we should really have a ppa that just has depends in it and does't have any of the nova packages in it... that way you don't have to worry about it | 23:09 |
| vish1 | mtaylor: which will have hot fixes that haven't been reviewed | 23:09 |
| *** Podilarius has left #openstack | 23:09 | |
| vish1 | i probably won't reference the public ppa | 23:09 |
| vish1 | i just want to make merging easy | 23:10 |
| mtaylor | oh, well, in that case, what I'd do is: | 23:10 |
| vish1 | so i'll probably have extra changelog entries for example | 23:10 |
| vish1 | and then i need to somehow handle the case when the patch makes it into the upstream | 23:10 |
| mtaylor | right. so I'd make a branch of lp:~nova-core/nova/debian-packaging - and a branch of lp:nova | 23:11 |
| mtaylor | make your hot-fix changes in lp:nova - so that you can push them back out for merging | 23:11 |
| mtaylor | and then when it's time to make packages, just go into your debian-packaging branch, do bzr merge ../hotfix-branch | 23:11 |
| mtaylor | and make whatever changelog entries you might need to make | 23:12 |
| mtaylor | whenever your hotfix branch gets merged, then it shoud all just magically work itself out | 23:12 |
| vish1 | ah ok | 23:13 |
| mtaylor | and bzr knows how to merge debian/changelog files - so it should do the right thing when new code has been pushed to lp:~nova-core/nova/debian-packaging | 23:13 |
| mtaylor | does that make any sort of sense? | 23:13 |
| vish1 | so i'll have to hand edit changelog still? | 23:13 |
| mtaylor | well, you can use dch to edit it | 23:13 |
| *** miclorb_ has quit IRC | 23:13 | |
| vish1 | hmm i'm getting errors in current trunk | 23:13 |
| vish1 | odd | 23:13 |
| mtaylor | yeah - the thing I just did wasn't for debian/changelog - it was to ship a root ChangeLog | 23:13 |
| *** miclorb has joined #openstack | 23:14 | |
| vish1 | ah ok | 23:14 |
| mtaylor | I think we need to figure out what's different on your machine from the hudson machine - yours seems to be stricter | 23:14 |
| vish1 | investigating build errors | 23:14 |
| vish1 | well i am on a mac | 23:14 |
| vish1 | :) | 23:14 |
| mtaylor | ah. well there it is :) | 23:14 |
| vish1 | and i may have different versions of packages installed | 23:14 |
| *** aliguori has quit IRC | 23:15 | |
| vish1 | oh i bet i know what it is | 23:16 |
| vish1 | hmm nm | 23:16 |
| vish1 | still looking | 23:16 |
| *** BartVB has joined #openstack | 23:18 | |
| *** miclorb has quit IRC | 23:18 | |
| vish1 | ok found the error | 23:21 |
| vish1 | don't have a symbolic link in the bin dir to the parent nova | 23:21 |
| vish1 | messes up some default relative paths | 23:21 |
| *** pvo has joined #openstack | 23:22 | |
| *** pvo has joined #openstack | 23:22 | |
| *** ChanServ sets mode: +v pvo | 23:22 | |
| *** rnewson has quit IRC | 23:23 | |
| mtaylor | heh | 23:24 |
| *** tr3buchet has quit IRC | 23:26 | |
| *** aliguori has joined #openstack | 23:29 | |
| eday | yay, swift ml created! | 23:34 |
| tjyang | Hi, all. After reading the 5 minutes overview and openstack FAQ and I still a have question. | 23:34 |
| mtaylor | eday: w00t | 23:35 |
| tjyang | this is my beginner question. Is openstack only good for Linux Cloud ? Can I use it on Solaris or HP-UX ? | 23:35 |
| *** pvo has quit IRC | 23:35 | |
| eday | tjyang: it's primarily being developed for linux at the moment, but there is no reason it could not work with hypervisors on other os's (assuming they have a sane Python environment) | 23:36 |
| vish1 | so i'm checking the merge rejection on auth-init | 23:36 |
| vish1 | i get no test fail | 23:36 |
| *** PiotrSikora has quit IRC | 23:37 | |
| tjyang | eday: thanks for the reply. My workplace have tons of Solaris machines at work. | 23:37 |
| vish1 | is it possible that the test run timed out? | 23:38 |
| vish1 | or that there was another test running at the same time? | 23:39 |
| mtaylor | no to another test running at same time | 23:39 |
| mtaylor | the box is serialized | 23:39 |
| mtaylor | I get occasional errors on that on my local box, re-running usually fixes | 23:39 |
| *** PiotrSikora has joined #openstack | 23:39 | |
| vish1 | hmm that is not good | 23:39 |
| vish1 | ok i'll try reproposing | 23:39 |
| mtaylor | just bump it back to Approved ... that should requeue it | 23:40 |
| tjyang | eday: so I may be able to build a solaris 10 cloud using LDOM hypervisor software ? | 23:41 |
| tjyang | eday: with openstack software. | 23:41 |
| vish1 | mtaylor: is there a standard way for adding things to python path for dev? | 23:41 |
| mtaylor | uh | 23:42 |
| vish1 | (or anyone else that knows) | 23:42 |
| *** miclorb has joined #openstack | 23:42 | |
| vish1 | so if i clone a repo, it isn't in the python path | 23:42 |
| mtaylor | I usually just make a virtualenv and install stuff in it - or install the stuff directly - although nova should support doing 'python setup.py develop' now | 23:42 |
| mtaylor | ah - perhaps what you're wanting is python setup.py develop | 23:43 |
| vish1 | we've been getting around it by adding a nova.pth | 23:43 |
| mtaylor | which will put .pth links in your system location for you | 23:43 |
| vish1 | to the dist-packages dir | 23:43 |
| mtaylor | yeah - develop will do that for you - one of the nice bits about having moved to setuptools | 23:43 |
| vish1 | yeah but the issue with that is that bzr stores branches in different dirs | 23:43 |
| mtaylor | ah. I see your issue | 23:43 |
| vish1 | so you just run setup.py develop every time you switch branches? | 23:43 |
| mtaylor | potentially. ... should come up with a better story for this | 23:44 |
| mtaylor | vish1: if you are in the root of the source tree, it should see the nova package because . is usually in the pythonpath, no? | 23:44 |
| vish1 | true | 23:45 |
| vish1 | but i'm not always in the root | 23:45 |
| mtaylor | gotcha | 23:45 |
| vish1 | specifically when running the bins | 23:45 |
| vish1 | they are not in the root | 23:45 |
| mtaylor | hrm. no. I do not have a good answer for you - but I will find one | 23:45 |
| vish1 | hence my magic with the symlink which unfortunately messes up paths | 23:45 |
| eday | tjyang: if there is a clean way to plug in solaris hypervisor modules into openstack, I think it's possible. RIght now it uses libvirt, so you would want to look at how that works. It would be interesting to see how well things line up | 23:46 |
| StylusEater | +creiht: submitted the merge proposal | 23:46 |
| tjyang | eday: I think I will spend some time to check openstatck (compute) out to get a feel. I don't want to see our internal IT machines got outsourced to external cloud vendor. | 23:49 |
| polvi | hey guys, anyone coming to the meet-up tomorrow? | 23:50 |
| tjyang | eday: our management show us http://www.salesforce.com/cloudcomputing/ video in one of our meeting :-< | 23:51 |
| vish1 | meet up at cloudkick? | 23:51 |
| polvi | vish1: yup | 23:51 |
| vish1 | polvi: i'll be there | 23:53 |
| vish1 | polvi: what time is it starting? | 23:53 |
| polvi | doors at 1pm, starting at 2pm sharp | 23:53 |
| *** silassewell has quit IRC | 23:56 | |
| *** sirp2 has quit IRC | 23:57 | |
| tjyang | I subscribed to openstack mailing list but the email address is <tmorey@rackspace.com>, is this correct ? | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!