Can you try again with the following command line options when you start
your client?

-Ddrill.exec.zk.retry.count=1

It looks like we have an artificially high attempt count of 7200 as our
default (with a 5 second timeout for each).

For reference, the settings are here:

https://github.com/apache/drill/blob/master/exec/java-exec/src/main/resources/drill-module.conf#L95



--
Jacques Nadeau
CTO and Co-Founder, Dremio

On Sun, Nov 1, 2015 at 1:36 PM, William Witt <[email protected]>
wrote:

> It does appear to be confined to zookeeper based connections.  I’ve
> expanded to a 3 node drill cluster, so connecting to a single drillbit
> seems innefficient.
>
> Will
>
> > On Oct 31, 2015, at 12:42 PM, Jacques Nadeau <[email protected]> wrote:
> >
> > I think this is a bug. Do you see the same behavior when you use a direct
> > connection string instead of a zookeeper based connection?
> > On Oct 29, 2015 9:39 AM, "William Witt" <[email protected]> wrote:
> >
> >> Im connecting to a drillbit via JDBC connection inside of a web app.  If
> >> the connection string is incorrect or if system can't connect for some
> >> reason it gets stuck in an infinite loop trying to connect.  Is this a
> bug
> >> am I doing something wrong?  Has anyone seen this before?
> >>
> >> Will
> >>
> >> Sent from my iPad
>
>

Reply via email to