> On 19 Oct 2014, at 17:18 , Benoit Chesneau <bchesn...@gmail.com> wrote: > > On Sun, Oct 12, 2014 at 10:50 AM, Alexander Shorin <kxe...@gmail.com> wrote: > >> Haven't tried to run release yet, but looking in dev/run script there >> are dozen config hacks to let cluster start. >> May be try this way for now? >> -- >> ,,,^..^,,, >> >> > Hum that's quite disappointing. I would expect that the release work in > 2.0 (and figured it would be tested). I guess that could be fixed though. > > Also what about running as a standalone node? I don't see any public > discussion on that topic but it seems that the consensus is to consider > that couchdb is now a cluster db: > > https://github.com/robertkowalski/couchdb/commit/2eaf50b16c4f0001a1fe1478330be2e77ce20896 > > which could explain why a release is not well tested or why it could hangs. > What is the position of other devs on this?
We haven’t gotten around to handle releases yet, any input welcome. So far I understand that we want to have CouchDB work in single node mode as simply as cluster mode. Best Jan -- > > - benoit > > > > >> >> On Sun, Oct 12, 2014 at 12:24 PM, Benoit Chesneau <bchesn...@gmail.com> >> wrote: >>> I dunno if I missed something after a succesful build, running couchdb >>> failed: >>> >>> https://www.friendpaste.com/7AjX8IxnwIQT11maMtD3f0 >>> >>> I guess it's a problem of starting the auth cache before having the user >> db >>> created but not sure. Before reporting it to jira, is there any step in >>> between before simply starting the release? >>> >>> - benoit >>
