You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@phoenix.apache.org by Ankit Singhal <an...@gmail.com> on 2016/07/25 15:28:06 UTC
Planning to cut RC by tomorrow morning IST
Moving PHOENIX-3111 to 4.8.1 as per discussion with Rajeshbabu. Let me know
if someone thinks it's important for 4.8.0.
Now currently, the only outstanding blocker for 4.8.0 is PHOENIX-3113 and
which has a simple fix. So, by tomorrow morning IST, we should have RC1 out.
Regards,
Ankit Singhal
On Mon, Jul 25, 2016 at 10:16 AM, James Taylor <ja...@apache.org>
wrote:
> Can we get an RC up today, please? There's one JIRA waiting to be committed
> and I think we're good after that. All licensing issues have been fixed.
> Thanks,
> James
>
> On Sunday, July 24, 2016, <la...@apache.org> wrote:
>
> > Also, can we branch 4.8-x in git?That way bigger changes can still go
> into
> > the 4.x branches, while bugs are fixed in 4.8-x branches.
> > -- Lars
> >
> > From: James Taylor <jamestaylor@apache.org <javascript:;>>
> > To: "dev@phoenix.apache.org <javascript:;>" <dev@phoenix.apache.org
> > <javascript:;>>
> > Sent: Thursday, July 21, 2016 7:16 PM
> > Subject: new 4.8.0 RC?
> >
> > How about a cutting a new RC now that the licensing work is complete,
> > Ankit? Looks like we can simplify the voting too by just having a single
> > vote across all versions by just include all the information in one VOTE
> > thread.
> >
> > Would be good to get PHOENIX-3078 in too, if possible. We've had a few
> > other fixes come in which IMHO are all ok to include in the new RC.
> >
> > Thanks,
> > James
> >
> >
> >
>
Re: Planning to cut RC by tomorrow morning IST
Posted by James Taylor <ja...@apache.org>.
Ok, thanks for the update, Rajeshbabu. My vote would be to hold off on the
RC until your fix is ready then.
On Mon, Jul 25, 2016 at 10:50 AM, rajeshbabu@apache.org <
chrajeshbabu32@gmail.com> wrote:
> It's happening frequently in our tests which include parallel create index,
> bulkload, deletes operations and automatic splits by HBase. Some times the
> dead lock happen with create index and some times with deletes.
>
> We already have a patch going under same tests. Mostly I can make final
> patch in 1 or 2 days.
>
> Thanks,
> Rajeshbabu.
>
> On Mon, Jul 25, 2016 at 9:48 PM, James Taylor <ja...@apache.org>
> wrote:
>
> > Thanks, Ankit. How likely is it that PHOENIX-3111 will happen? Since it's
> > not a regression and this is the first we've heard of it, maybe not very
> > likely? Or is it more common with the new local index implementation? If
> > it's a matter of a few days, probably worth holding off on the RC.
> >
> > Thoughts?
> >
> > On Mon, Jul 25, 2016 at 8:28 AM, Ankit Singhal <ankitsinghal59@gmail.com
> >
> > wrote:
> >
> > > Moving PHOENIX-3111 to 4.8.1 as per discussion with Rajeshbabu. Let me
> > know
> > > if someone thinks it's important for 4.8.0.
> > > Now currently, the only outstanding blocker for 4.8.0 is PHOENIX-3113
> and
> > > which has a simple fix. So, by tomorrow morning IST, we should have RC1
> > > out.
> > >
> > > Regards,
> > > Ankit Singhal
> > >
> > >
> > >
> > > On Mon, Jul 25, 2016 at 10:16 AM, James Taylor <jamestaylor@apache.org
> >
> > > wrote:
> > >
> > > > Can we get an RC up today, please? There's one JIRA waiting to be
> > > committed
> > > > and I think we're good after that. All licensing issues have been
> > fixed.
> > > > Thanks,
> > > > James
> > > >
> > > > On Sunday, July 24, 2016, <la...@apache.org> wrote:
> > > >
> > > > > Also, can we branch 4.8-x in git?That way bigger changes can still
> go
> > > > into
> > > > > the 4.x branches, while bugs are fixed in 4.8-x branches.
> > > > > -- Lars
> > > > >
> > > > > From: James Taylor <jamestaylor@apache.org <javascript:;>>
> > > > > To: "dev@phoenix.apache.org <javascript:;>" <
> dev@phoenix.apache.org
> > > > > <javascript:;>>
> > > > > Sent: Thursday, July 21, 2016 7:16 PM
> > > > > Subject: new 4.8.0 RC?
> > > > >
> > > > > How about a cutting a new RC now that the licensing work is
> complete,
> > > > > Ankit? Looks like we can simplify the voting too by just having a
> > > single
> > > > > vote across all versions by just include all the information in one
> > > VOTE
> > > > > thread.
> > > > >
> > > > > Would be good to get PHOENIX-3078 in too, if possible. We've had a
> > few
> > > > > other fixes come in which IMHO are all ok to include in the new RC.
> > > > >
> > > > > Thanks,
> > > > > James
> > > > >
> > > > >
> > > > >
> > > >
> > >
> >
>
Re: Planning to cut RC by tomorrow morning IST
Posted by "rajeshbabu@apache.org" <ch...@gmail.com>.
It's happening frequently in our tests which include parallel create index,
bulkload, deletes operations and automatic splits by HBase. Some times the
dead lock happen with create index and some times with deletes.
We already have a patch going under same tests. Mostly I can make final
patch in 1 or 2 days.
Thanks,
Rajeshbabu.
On Mon, Jul 25, 2016 at 9:48 PM, James Taylor <ja...@apache.org>
wrote:
> Thanks, Ankit. How likely is it that PHOENIX-3111 will happen? Since it's
> not a regression and this is the first we've heard of it, maybe not very
> likely? Or is it more common with the new local index implementation? If
> it's a matter of a few days, probably worth holding off on the RC.
>
> Thoughts?
>
> On Mon, Jul 25, 2016 at 8:28 AM, Ankit Singhal <an...@gmail.com>
> wrote:
>
> > Moving PHOENIX-3111 to 4.8.1 as per discussion with Rajeshbabu. Let me
> know
> > if someone thinks it's important for 4.8.0.
> > Now currently, the only outstanding blocker for 4.8.0 is PHOENIX-3113 and
> > which has a simple fix. So, by tomorrow morning IST, we should have RC1
> > out.
> >
> > Regards,
> > Ankit Singhal
> >
> >
> >
> > On Mon, Jul 25, 2016 at 10:16 AM, James Taylor <ja...@apache.org>
> > wrote:
> >
> > > Can we get an RC up today, please? There's one JIRA waiting to be
> > committed
> > > and I think we're good after that. All licensing issues have been
> fixed.
> > > Thanks,
> > > James
> > >
> > > On Sunday, July 24, 2016, <la...@apache.org> wrote:
> > >
> > > > Also, can we branch 4.8-x in git?That way bigger changes can still go
> > > into
> > > > the 4.x branches, while bugs are fixed in 4.8-x branches.
> > > > -- Lars
> > > >
> > > > From: James Taylor <jamestaylor@apache.org <javascript:;>>
> > > > To: "dev@phoenix.apache.org <javascript:;>" <dev@phoenix.apache.org
> > > > <javascript:;>>
> > > > Sent: Thursday, July 21, 2016 7:16 PM
> > > > Subject: new 4.8.0 RC?
> > > >
> > > > How about a cutting a new RC now that the licensing work is complete,
> > > > Ankit? Looks like we can simplify the voting too by just having a
> > single
> > > > vote across all versions by just include all the information in one
> > VOTE
> > > > thread.
> > > >
> > > > Would be good to get PHOENIX-3078 in too, if possible. We've had a
> few
> > > > other fixes come in which IMHO are all ok to include in the new RC.
> > > >
> > > > Thanks,
> > > > James
> > > >
> > > >
> > > >
> > >
> >
>
Re: Planning to cut RC by tomorrow morning IST
Posted by James Taylor <ja...@apache.org>.
Thanks, Ankit. How likely is it that PHOENIX-3111 will happen? Since it's
not a regression and this is the first we've heard of it, maybe not very
likely? Or is it more common with the new local index implementation? If
it's a matter of a few days, probably worth holding off on the RC.
Thoughts?
On Mon, Jul 25, 2016 at 8:28 AM, Ankit Singhal <an...@gmail.com>
wrote:
> Moving PHOENIX-3111 to 4.8.1 as per discussion with Rajeshbabu. Let me know
> if someone thinks it's important for 4.8.0.
> Now currently, the only outstanding blocker for 4.8.0 is PHOENIX-3113 and
> which has a simple fix. So, by tomorrow morning IST, we should have RC1
> out.
>
> Regards,
> Ankit Singhal
>
>
>
> On Mon, Jul 25, 2016 at 10:16 AM, James Taylor <ja...@apache.org>
> wrote:
>
> > Can we get an RC up today, please? There's one JIRA waiting to be
> committed
> > and I think we're good after that. All licensing issues have been fixed.
> > Thanks,
> > James
> >
> > On Sunday, July 24, 2016, <la...@apache.org> wrote:
> >
> > > Also, can we branch 4.8-x in git?That way bigger changes can still go
> > into
> > > the 4.x branches, while bugs are fixed in 4.8-x branches.
> > > -- Lars
> > >
> > > From: James Taylor <jamestaylor@apache.org <javascript:;>>
> > > To: "dev@phoenix.apache.org <javascript:;>" <dev@phoenix.apache.org
> > > <javascript:;>>
> > > Sent: Thursday, July 21, 2016 7:16 PM
> > > Subject: new 4.8.0 RC?
> > >
> > > How about a cutting a new RC now that the licensing work is complete,
> > > Ankit? Looks like we can simplify the voting too by just having a
> single
> > > vote across all versions by just include all the information in one
> VOTE
> > > thread.
> > >
> > > Would be good to get PHOENIX-3078 in too, if possible. We've had a few
> > > other fixes come in which IMHO are all ok to include in the new RC.
> > >
> > > Thanks,
> > > James
> > >
> > >
> > >
> >
>