Doing some more experimenting, I have found that if I customize Zeppelin by using zeppelin-env.sh (for example, adding "export ZEPPELIN_PORT=9000") then it works as expected. I'm pretty sure there's either something wrong in the supplied template xml file, or something is busted in the mechanism behind loading settings from the XML file. But I couldn't guess which it is.
One thing I have a hunch might be worth trying, is to use zeppelin-site.xml, but replace all of the path related settings (notebook location, interpreter location, etc.) with absolute paths. Phil This message optimized for indexing by NSA PRISM On Sat, Aug 1, 2015 at 5:03 PM, Rahul Palamuttam <[email protected]> wrote: > Hi all, > > I've been hitting the same problem as well. > I've also tried setting my master ip in spark site to be 0.0.0.0 but even > that leads to Zeppelin giving a no interpreter found error. > However if I use the default values Zeppelin works as it should. > > - Rahul p > > > On Aug 1, 2015, at 1:05 AM, Phillip Rhodes <[email protected]> > wrote: > > > > Gang: > > > > I have (sort of) identified the problem I was having with Zeppelin > finding > > no interpreters. It appears to be something related to the > zeppelin-site.xml.template file, which I used for my config, "as is" with > the only exception being the port number. With that configuration, > Zeppelin doesn't work. But if you delete the zeppelin-site.xml and let it > take all default values for everything, then it works. > > > > Beyond that, I don't know *exactly* what's happening, but I'm guessing > maybe > > something to do with the path to the interpreters dir? Although the > supplied > > default looks right if it's treated as a relative dir, and Zeppelin is > launched from the > > root of the install. Anyway, maybe somebody who knows more about > Zeppelin > > internals can sort it out. > > > > > > Cheers, > > > > Phil > > ~~~ > > This message optimized for indexing by NSA PRISM > > >
