December 11, 2015 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
December 11, 2015 [00:17:27] *** pamaury has quit IRC[00:18:35] *** pamaury has joined #gerrit[00:39:57] *** gbeardshaw has joined #gerrit[00:42:59] <gbeardshaw> Hi guys - could anybody tell me any major pitfalls they can think of in upgrading from Gerrit v2.1.6.1 to 2.11.5 or point towards some good upgrade documentation? I can't seem to find much / anything related to upgrades in the main documentation.[00:43:10] <gbeardshaw> *please* :-)[00:52:56] *** qchris has quit IRC[00:57:46] <gbeardshaw> hmmm, looks like I need to jump through 2.1.7 first, has anybody got any advice of any other release I'd need to make to get to the latest release? Given the small jump between 2.1.6 and 2.1.7 I'd expect there's likely other jumps I'd need to do? Basically, I guess I'm just looking for the easiest and quickest way to jump between 2.1.6.1 to 2.11.5..[00:59:34] <sbeller> depends if you can take your system offline or if you want to do online reindexing. If you want to do it online, you need to go version by version[01:00:16] <sbeller> if you take it offline, you *could* skip over and just go to the latest version and rebuild the database IIUC (never done that myself)[01:02:01] <gbeardshaw> I can take it offline. IIUC?[01:03:04] <gbeardshaw> I definitely need to preserve the data, though.[01:04:12] <gbeardshaw> Would I just run the standard "gerrit.war init" command though or would I need --batch too?[01:05:04] <gbeardshaw> I think I'm definitely doing to have to test this but I guess I just need some steer from people in the "know" :-)[01:05:21] <gbeardshaw> going*[02:29:22] *** pamaury has quit IRC[03:01:28] <gbeardshaw> Renaming "-- All Projects --" to "All-Projects"[03:01:34] <gbeardshaw> Exception in thread "main" java.lang.NullPointerException[03:01:43] <gbeardshaw> Any idea what that's about anybody?[03:11:31] *** gbeardshaw has quit IRC[03:30:37] *** Vampire0_ has joined #gerrit[03:33:46] *** Vampire0 has quit IRC[06:26:01] *** Tica2 has joined #gerrit[07:24:34] *** ch007m has joined #gerrit[07:27:20] *** yena has quit IRC[07:27:58] *** yena has joined #gerrit[07:37:15] *** lqi has joined #gerrit[07:39:35] *** ch007m has quit IRC[07:39:40] *** ch007m_2 has joined #gerrit[07:52:00] *** ch007m_2 has quit IRC[07:53:39] *** ch007m has joined #gerrit[07:56:25] *** ch007m_2 has joined #gerrit[07:58:06] *** ch007m has quit IRC[09:24:10] *** HeOS has quit IRC[09:47:29] *** Haza`Aw is now known as Haza`[10:11:00] *** qchris has joined #gerrit[10:16:44] *** HeOS has joined #gerrit[11:07:04] *** firemanxbr has joined #gerrit[11:14:56] *** lqi has quit IRC[11:43:50] *** lqi has joined #gerrit[12:15:27] *** plinio has joined #gerrit[12:16:52] *** farosas_ has joined #gerrit[12:27:07] *** pamaury has joined #gerrit[12:36:21] *** Haza` is now known as Haza`Aw[12:41:06] *** pamaury has quit IRC[13:26:11] *** farosas_ has quit IRC[13:42:27] *** farosas_ has joined #gerrit[13:59:09] *** Haza`Aw is now known as Haza`[14:39:56] *** HeOS has quit IRC[14:42:39] *** Haza` is now known as Haza`Aw[15:29:49] *** peristeri has joined #gerrit[15:31:06] *** Haza`Aw is now known as Haza`[15:55:54] *** Haza` is now known as Haza`Aw[15:58:05] *** mafi has left #gerrit[16:12:28] *** dcaro has joined #gerrit[16:14:09] <dcaro> Hi everyone, I have a question of how to work with feature branches on gerrit, the issue is that when I push a merge commit to review to the master branch from the feature one, it creates one patch-set for each of the branch commits, and that allows mergine the commits from the feature branch separatedly as opposed to merging them together with the merge commit[16:15:01] <dcaro> am I doing anything wrong? is that a short comming of gerrit? I can create a single squashed commit instead of a merge one, but that would not keep it's history[16:18:55] <alkino> there is no branch "idea" on gerrit.[16:19:03] <alkino> AFAIK[16:20:32] <VLetrmx> dcaro, the closest thing to a feature branch in gerrit is a set of changes with the same topic pretty much[16:21:45] <VLetrmx> so yes currently you have to merge each change individually, though there will i believe be a way to merge an entire topic in the future[16:24:20] <dcaro> VLetrmx: mmm... that would be nice, do you know if there's a ticket I can follow?[16:24:43] <jumpula> if you don't rebase your feature branch, isn't it possible just to merge that in master, push it for review and you'll just get a review for the merge commit?[16:25:14] <dcaro> jumpula: you get a new change for each of the commits in the feature branch + one for the merge commit[16:25:30] <jumpula> however, if you've rebased reviews that are closed, i think it'll open a new review for each of those commits in the merge commit[16:25:45] <jumpula> ie. the question is, did you rebase your feature branch?[16:26:01] <dcaro> jumpula: yes and no, tried both, same result[16:26:29] <jumpula> interesting[16:26:52] <alkino> dcaro: if we have a big feature branch, we review it on gitlab and we push it to gerrit without review.[16:26:54] <dcaro> jumpula: you just get one change for each commit in the feature brach (with target branch the master one) + one change for the merge commit (with two parents)[16:27:14] <dcaro> alkino: :/[16:27:15] <alkino> git push gerrit HEAD:/refs/for/master[16:27:30] <alkino> we have a gitlab synchronized on gerrit/master[16:27:32] <dcaro> alkino: but that will open one change for each commit no?[16:27:37] <alkino> no ;)[16:27:43] <dcaro> how come?[16:27:45] <jumpula> no, because it bypasses the review altogether[16:27:53] <jumpula> it's like pushing directly in git[16:28:05] <dcaro> to bypass review you have to push to HEAD:refs/heads/master no?[16:28:19] <jumpula> ah, yeah[16:28:30] <alkino> yeah[16:28:32] <alkino> sorry[16:28:32] <jumpula> refs/for/branch is review[16:28:33] <dcaro> okok[16:28:54] <jumpula> somehow i read alkinos line as refs/heads/master[16:28:58] <dcaro> not nice to have to handle two review system though[16:29:06] <jumpula> but i assume it's a typo. the first / is a typo, at least.[16:29:22] <jumpula> well, you give a certain group access to push directly in master[16:29:27] <jumpula> and you trust that group[16:29:31] <alkino> ra git push gerrit HEAD:refs/heads/master[16:29:39] <jumpula> if merges are trivial, they're not interesting review material anyway[16:30:01] <jumpula> and if you rebase your feature branch pre-merge, they are trivial[16:30:35] <dcaro> yes, issue is that you have to maintain a gitlab too (or github) and run ci there too, and reviews[16:31:02] <dcaro> so essentially, just using gerrit as mirror mostly iiuc[16:32:01] <alkino> no[16:32:16] <dcaro> good, please elaborate :_)[16:32:26] <alkino> in fact yes[16:32:46] <jumpula> :D[16:32:48] <dcaro> hehehe[16:33:00] <alkino> we don't do it. We push to gerrit, review on gitlab, run jenkins on gerrit. And push the branch to gerrit. Abandon all patches on gerrit.[16:33:03] <alkino> it's not really clean[16:34:01] <alkino> we work around the lack of this feature.[16:34:05] <dcaro> yep[16:34:13] <jumpula> i'm thinking everyone does that[16:34:17] <dcaro> hahaha[16:34:23] <dcaro> why not ask for the feature?[16:34:49] <jumpula> i'm relatively new to gerrit, after two weeks of trying things out, i haven't figured out a clean way to merge feature branches back in mainline[16:35:19] <jumpula> beside directly pushing merged branches in, without review[16:36:32] <VLetrmx> dcaro, it's in master but experimental and disabled by default ( https://gerrit-documentation.storage.googleapis.com/Documentation/2.12/config-gerrit.html#change.submitWholeTopic )[16:36:44] <dcaro> nice!![16:39:01] <jumpula> but that assumes you've pushed things to refs/for/master?[16:39:09] <jumpula> and you have topic set for each change[16:39:19] <dcaro> yep[16:39:42] <dcaro> git review helps adding a topic to the change when sending for review[16:39:51] <jumpula> so in effect, if you've worked in a feature branch, pushed stuff to refs/for/feature, how does it help?[16:40:19] <dcaro> so you'd push to refs/for/master/feature[16:40:24] <jumpula> dcaro: what about when you incrementally do patches within a branch and don't push them to review all at once?[16:40:30] <dcaro> that will add the 'feature' topic to all the patches[16:40:40] <jumpula> ah, yes.[16:40:59] <dcaro> then you just hit 'submit' and it will merge all of them at once[16:41:17] <dcaro> (you could add a merge commit, and you'll have the nice history too)[16:42:07] <dcaro> mmm... now I want that feature in :)[16:43:00] <jumpula> but not being able to merge across branches is still a problem[16:43:12] <jumpula> say you want to try out our feature in staging and release branches[16:43:20] <dcaro> you can merge, just make sure they go to the same topic[16:43:33] <jumpula> or your feature didn't make it to release branch, but you've branched the topic from there[16:44:13] <jumpula> ie. merging refs/for/master/feature -> refs/for/staging[16:44:31] <jumpula> or something like that. i suspect there will be a review for all of the commits... again.[16:45:25] <dcaro> yep, you'll get a change for each commit[16:45:44] <jumpula> so that support is incomplete for some workflows[16:46:04] <dcaro> agree[16:47:00] *** lqi has quit IRC[16:47:42] *** Tica2 has quit IRC[16:48:45] *** lqi has joined #gerrit[16:51:40] *** lqi has joined #gerrit[17:10:32] *** mafi has joined #gerrit[17:39:44] *** lqi has quit IRC[17:40:18] *** Tica2 has joined #gerrit[17:42:15] *** lqi has joined #gerrit[18:22:24] *** firemanxbr has quit IRC[18:47:35] *** Tica2 has quit IRC[18:53:23] *** Tica2 has joined #gerrit[19:26:34] *** ch007m_2 has quit IRC[19:28:51] *** lqi has quit IRC[19:29:27] *** lqi has joined #gerrit[19:34:35] *** twk has joined #gerrit[19:34:58] *** twk has joined #gerrit[19:34:58] *** twk has quit IRC[19:34:58] *** twk has joined #gerrit[19:37:03] *** Tica2 has quit IRC[19:37:40] *** Tica2 has joined #gerrit[19:50:05] *** Tica2 has quit IRC[19:56:25] *** Tica2 has joined #gerrit[19:57:51] *** Tica2 has quit IRC[19:58:18] *** Tica2 has joined #gerrit[19:59:20] *** lqi has quit IRC[20:49:16] *** pamaury has joined #gerrit[21:00:20] *** plinio has quit IRC[21:01:11] *** alkino has quit IRC[21:08:33] *** farosas_ has quit IRC[21:10:27] *** lqi has joined #gerrit[22:04:04] *** peristeri has quit IRC[22:04:24] *** peristeri has joined #gerrit[22:09:09] *** Tica2 has quit IRC[22:15:02] *** peristeri has quit IRC[22:16:19] *** Tica2 has joined #gerrit[22:43:05] *** Tica2 has quit IRC[22:43:46] *** Tica2 has joined #gerrit[23:03:11] *** dcaro is now known as dcaro_nothere[23:52:31] *** Puhe has joined #gerrit[23:58:29] *** jrnieder has quit IRC