Yes, that's what I just found... I will have to use a persistent volume to store the config file. This should be redesigned, by making changes to that and other files... it's really hard to dockerize and deploy Zeppelin to kubernetes, mesos, ... without having to resort to bootstrapping scripts and other magic. Regards,
Luis Angel Vicente Sanchez zeppelin-us...@bigcente.ch On Thu, 31 Aug 2017, at 17:56, moon soo Lee wrote: > Thanks for sharing the idea. > Sounds very interesting and useful. > > I wish this feature be general enough so we can use it with not only > JDBC but also with SparkSQL, Pandas dataframe, etc.> > Then i guess internal data representation will be important topic. > Zeppelin community had several attempt and related discussions. > > Following proposal is one of proposal that Zeppelin community made. > https://cwiki.apache.org/confluence/display/ZEPPELIN/Proposal%3A+Extending+TableData+API> > This proposal talk more about foundation of internal data > representation.> But i believe it opens many possibilities like metadata > browser, cross > interpreter join, etc.> > What do you think? > > Thanks, > moon > > On Tue, Aug 29, 2017 at 2:38 AM Polyakov Valeriy > <v.polja...@tinkoff.ru> wrote:>> Hello, Team!____ >> __ __ >> What do you think about implementation of high-grade metadata browser >> for JDBC interpreter? As it is done in any popular SQL development >> tools, for example, DataGrip. I'm talking about new detailed panel >> which will contain metadata info of all included in note >> interpreters. I've found only one task [1] with a similar problem, >> but in my opinion its objectives are too limited. We could work on >> something more ambitious and convenient. So, I would like to hear >> your opinion on this matter.____>> __ __ >> Thank you!____ >> __ __ >> [1] https://issues.apache.org/jira/browse/ZEPPELIN-1824____ >> __ __ >> __ __ >> __ __ >> Kind regards, >> *Valeriy Polyakov*____ >> __ __
