metron-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Casey Stella <ceste...@gmail.com>
Subject Re: Kerberos changes affected quick-dev and full-dev
Date Mon, 03 Apr 2017 18:33:08 GMT
Thanks Justin,  the packer build is started, but this is going to take some
time.  Please use full-dev to validate your PRs in the meantime.  I will
update this thread once it's uploaded.

On Mon, Apr 3, 2017 at 2:13 PM, Justin Leet <justinjleet@gmail.com> wrote:

> The PR to fix full-dev is in master now.  We still need a new packer build
> before we have quick-dev available again.
>
> Justin
>
> On Mon, Apr 3, 2017 at 10:53 AM, Justin Leet <justinjleet@gmail.com>
> wrote:
>
> > Btw, here is a workaround for full-dev. In Ambari, add the line
> "topology.worker.childopts="
> > (no argument) to the elasticsearch.properties template, then restart
> > indexing through Ambari to propogate the change out.
> >
> > For example, make the Storm section look like:
> >
> > ##### Storm #####
> > indexing.workers=1
> > indexing.executors=0
> > topology.worker.childopts=
> >
> > Justin
> >
> > On Mon, Apr 3, 2017 at 10:46 AM, Casey Stella <cestella@gmail.com>
> wrote:
> >
> >> Hey guys,
> >>
> >> Just a quick heads up, the kerberos related changes (797 and 793) that
> >> went
> >> in last week had mpack changes.  This means that a new packer build
> needs
> >> to be updated for quickdev to work.  Unfortunately, that didn't happen
> >> *and* there's a follow-on bug (METRON-818) that also involves mpack
> >> changes
> >> (https://github.com/apache/incubator-metron/pull/506).
> >>
> >> Also, this change fixes a bug introduced in 797 with full-dev, so it's
> >> getting high priority attention just right now.  I just wanted to send
> an
> >> update and make sure everyone was aware of what's going on if you try
> >> full-dev or quick-dev and it fails for you.  I expect 818 to get in
> >> quickly
> >> (already has 2 +1s, so pretty much as soon as travis returns we'll
> >> commit),
> >> which should fix full-dev.
> >>
> >> Casey
> >>
> >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message