Re: [CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[rajeshb...@apache.org]

Thanks Istvan, Richard, Cong Luo for fixing the issues quickly.


Will create another RC today.

Thanks,
Rajeshbabu.


On Wed, Jan 24, 2024, 12:34 PM IstvaToth  wrote:

> Every problem found in  RC0 should be fixed now.
> (Plus I have also merged OMID-275)
>
> Istvan
>
> On Tue, Jan 23, 2024 at 8:34 AM Istvan Toth  wrote:
>
> > It would also be nice if OMID-275
> >  could make it into
> 1.1.1.
> >
> > It's a small change, but it would open the possibility of building
> > 5.1.4 with future Omid releases which will have to remove
> > TTable.getHgetTableDescriptor() for HBase 3 compatibility.
> >
> > best regards
> > Istvan
> >
> >
> > On Fri, Jan 19, 2024 at 2:58 PM Istvan Toth  wrote:
> >
> >> We have worked on this with Richard.
> >>
> >> I have made some changes to force the test to use IPv6 on my Linux
> >> machine, and it still works fine.
> >> It is some funky mac networking issue:
> >>
> >>
> https://medium.com/@quelgar/java-sockets-broken-for-ipv6-on-mac-5aae72f06b21
> >>
> >> Please review https://issues.apache.org/jira/browse/OMID-249, which
> >> fixes the issue (at least on Richard's particular mac)
> >>
> >> best regards
> >> Istvan
> >>
> >> On Fri, Jan 19, 2024 at 8:07 AM Istvan Toth  wrote:
> >>
> >>> Someone who has access to a Mac please try to track down which change
> >>> broke the tests.
> >>>
> >>>
> >>> On Fri, Jan 19, 2024 at 8:05 AM Istvan Toth 
> wrote:
> >>>
>  PR for JDK17 support is up:
>  https://issues.apache.org/jira/browse/OMID-272
> 
>  On Thu, Jan 18, 2024 at 6:39 PM rajeshb...@apache.org <
>  chrajeshbab...@gmail.com> wrote:
> 
> > The vote has been canceled to address issues with JDK and Ipv6
> > interface in
> > MacOS
> >
> > Thanks,
> > Rajeshbabu.
> >
> > On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
> > chrajeshbab...@gmail.com> wrote:
> >
> > > Sure, let me cancel the RC and create another RC after fixes with
> the
> > > network interface and JDK 17.
> > >
> > > Thanks, Richard and Istvan for finding issues, and Cong Luo for
> your
> > > support.
> > >
> > >
> > >
> > >
> > >
> > > On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth
> > 
> > > wrote:
> > >
> > >> I have found another problem, Omid doesn't work with JDK 17.
> > >> Not necessarily a blocker, but if we do another RC, it would be
> > nice to
> > >> include it.
> > >>
> > >> I'm also working on adding HBase 3 support, which requires a small
> > API
> > >> change (that also affects Phoenix)
> > >> Again, not a blocker, but we'll have to do another release for 5.2
> > sooner
> > >> than later for HBase 3 support.
> > >>
> > >>
> > >> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
> > >>
> > >> > Hi Istvan,
> > >> > Using ipv6 on linux operating systems is available, so the
> reason
> > may be
> > >> > limited to users using Mac OS. And then, I using the ping the
> ipv6
> > >> address
> > >> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as
> > the host
> > >> > name. based on this default behavior of the operating system, I
> > >> recommend
> > >> > using macos with ipv4 and others with ipv4/ipv6.
> > >> >
> > >> > On 2024/01/18 06:54:14 Istvan Toth wrote:
> > >> > > I think we should look into this a bit deeper.
> > >> > > Why does this fail with IPv6 ?
> > >> > >
> > >> > > Are we explicitly binding TSO to the IPv4 address only ?
> > >> > > Do we only do that in the tests, or does that happen in
> > production ?
> > >> > > Is it even possible to use Omid with IPv6 ?
> > >> > >
> > >> > > I suspect that fixing
> > https://issues.apache.org/jira/browse/OMID-249
> > >> > would
> > >> > > improve on this.
> > >> > >
> > >> > > Generally, I think that having a single IP per TSO can be a
> > problem
> > >> in a
> > >> > > mixed IPv4 / IPv6 network, and with some network
> configurations.
> > >> > > Maybe we should store the name of the active server instead of
> > >> addresses
> > >> > in
> > >> > > ZK, and kick this problem down to the network/DNS
> administrator.
> > >> > >
> > >> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK
> > with
> > >> Cong's
> > >> > > proposed solution, but if it breaks pre-existing Ipv6 support,
> > then we
> > >> > > should
> > >> > > fix this properly for 1.1.1.
> > >> > >
> > >> > > We should create a follow-up ticket to make sure the Omid is
> > IPv6
> > >> clean
> > >> > for
> > >> > > the next release at the very least.
> > >> > >
> > >> > > Istvan
> > >> > >
> > >> > >
> > >> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo 
> > wrote:
> > >> > >
> > >> > > > Hi Richard,
> > >> > > >   Actually, I 

Re: [CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[Istvan Toth]

Every problem found in  RC0 should be fixed now.
(Plus I have also merged OMID-275)

Istvan

On Tue, Jan 23, 2024 at 8:34 AM Istvan Toth  wrote:

> It would also be nice if OMID-275
>  could make it into 1.1.1.
>
> It's a small change, but it would open the possibility of building
> 5.1.4 with future Omid releases which will have to remove
> TTable.getHgetTableDescriptor() for HBase 3 compatibility.
>
> best regards
> Istvan
>
>
> On Fri, Jan 19, 2024 at 2:58 PM Istvan Toth  wrote:
>
>> We have worked on this with Richard.
>>
>> I have made some changes to force the test to use IPv6 on my Linux
>> machine, and it still works fine.
>> It is some funky mac networking issue:
>>
>> https://medium.com/@quelgar/java-sockets-broken-for-ipv6-on-mac-5aae72f06b21
>>
>> Please review https://issues.apache.org/jira/browse/OMID-249, which
>> fixes the issue (at least on Richard's particular mac)
>>
>> best regards
>> Istvan
>>
>> On Fri, Jan 19, 2024 at 8:07 AM Istvan Toth  wrote:
>>
>>> Someone who has access to a Mac please try to track down which change
>>> broke the tests.
>>>
>>>
>>> On Fri, Jan 19, 2024 at 8:05 AM Istvan Toth  wrote:
>>>
 PR for JDK17 support is up:
 https://issues.apache.org/jira/browse/OMID-272

 On Thu, Jan 18, 2024 at 6:39 PM rajeshb...@apache.org <
 chrajeshbab...@gmail.com> wrote:

> The vote has been canceled to address issues with JDK and Ipv6
> interface in
> MacOS
>
> Thanks,
> Rajeshbabu.
>
> On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
> chrajeshbab...@gmail.com> wrote:
>
> > Sure, let me cancel the RC and create another RC after fixes with the
> > network interface and JDK 17.
> >
> > Thanks, Richard and Istvan for finding issues, and Cong Luo for your
> > support.
> >
> >
> >
> >
> >
> > On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth
> 
> > wrote:
> >
> >> I have found another problem, Omid doesn't work with JDK 17.
> >> Not necessarily a blocker, but if we do another RC, it would be
> nice to
> >> include it.
> >>
> >> I'm also working on adding HBase 3 support, which requires a small
> API
> >> change (that also affects Phoenix)
> >> Again, not a blocker, but we'll have to do another release for 5.2
> sooner
> >> than later for HBase 3 support.
> >>
> >>
> >> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
> >>
> >> > Hi Istvan,
> >> > Using ipv6 on linux operating systems is available, so the reason
> may be
> >> > limited to users using Mac OS. And then, I using the ping the ipv6
> >> address
> >> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as
> the host
> >> > name. based on this default behavior of the operating system, I
> >> recommend
> >> > using macos with ipv4 and others with ipv4/ipv6.
> >> >
> >> > On 2024/01/18 06:54:14 Istvan Toth wrote:
> >> > > I think we should look into this a bit deeper.
> >> > > Why does this fail with IPv6 ?
> >> > >
> >> > > Are we explicitly binding TSO to the IPv4 address only ?
> >> > > Do we only do that in the tests, or does that happen in
> production ?
> >> > > Is it even possible to use Omid with IPv6 ?
> >> > >
> >> > > I suspect that fixing
> https://issues.apache.org/jira/browse/OMID-249
> >> > would
> >> > > improve on this.
> >> > >
> >> > > Generally, I think that having a single IP per TSO can be a
> problem
> >> in a
> >> > > mixed IPv4 / IPv6 network, and with some network configurations.
> >> > > Maybe we should store the name of the active server instead of
> >> addresses
> >> > in
> >> > > ZK, and kick this problem down to the network/DNS administrator.
> >> > >
> >> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK
> with
> >> Cong's
> >> > > proposed solution, but if it breaks pre-existing Ipv6 support,
> then we
> >> > > should
> >> > > fix this properly for 1.1.1.
> >> > >
> >> > > We should create a follow-up ticket to make sure the Omid is
> IPv6
> >> clean
> >> > for
> >> > > the next release at the very least.
> >> > >
> >> > > Istvan
> >> > >
> >> > >
> >> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo 
> wrote:
> >> > >
> >> > > > Hi Richard,
> >> > > >   Actually, I had the same issue before, but it was resolved
> >> locally:
> >> > if
> >> > > > the NIC was assigned both ipv4 and ipv6 addresses, use ipv4
> first.
> >> If
> >> > we
> >> > > > all think this is acceptable, then I will create a PR
> submission.
> >> > > >
> >> > > > On 2024/01/17 14:30:50 Richárd Antal wrote:
> >> > > > > -1
> >> > > > >
> >> > > > > Signature: OK
> >> > > > > Checksum: OK
> >> > > > > m

Re: [CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[Istvan Toth]

It would also be nice if OMID-275
 could make it into 1.1.1.

It's a small change, but it would open the possibility of building
5.1.4 with future Omid releases which will have to remove
TTable.getHgetTableDescriptor() for HBase 3 compatibility.

best regards
Istvan


On Fri, Jan 19, 2024 at 2:58 PM Istvan Toth  wrote:

> We have worked on this with Richard.
>
> I have made some changes to force the test to use IPv6 on my Linux
> machine, and it still works fine.
> It is some funky mac networking issue:
>
> https://medium.com/@quelgar/java-sockets-broken-for-ipv6-on-mac-5aae72f06b21
>
> Please review https://issues.apache.org/jira/browse/OMID-249, which fixes
> the issue (at least on Richard's particular mac)
>
> best regards
> Istvan
>
> On Fri, Jan 19, 2024 at 8:07 AM Istvan Toth  wrote:
>
>> Someone who has access to a Mac please try to track down which change
>> broke the tests.
>>
>>
>> On Fri, Jan 19, 2024 at 8:05 AM Istvan Toth  wrote:
>>
>>> PR for JDK17 support is up:
>>> https://issues.apache.org/jira/browse/OMID-272
>>>
>>> On Thu, Jan 18, 2024 at 6:39 PM rajeshb...@apache.org <
>>> chrajeshbab...@gmail.com> wrote:
>>>
 The vote has been canceled to address issues with JDK and Ipv6
 interface in
 MacOS

 Thanks,
 Rajeshbabu.

 On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
 chrajeshbab...@gmail.com> wrote:

 > Sure, let me cancel the RC and create another RC after fixes with the
 > network interface and JDK 17.
 >
 > Thanks, Richard and Istvan for finding issues, and Cong Luo for your
 > support.
 >
 >
 >
 >
 >
 > On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth
 
 > wrote:
 >
 >> I have found another problem, Omid doesn't work with JDK 17.
 >> Not necessarily a blocker, but if we do another RC, it would be nice
 to
 >> include it.
 >>
 >> I'm also working on adding HBase 3 support, which requires a small
 API
 >> change (that also affects Phoenix)
 >> Again, not a blocker, but we'll have to do another release for 5.2
 sooner
 >> than later for HBase 3 support.
 >>
 >>
 >> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
 >>
 >> > Hi Istvan,
 >> > Using ipv6 on linux operating systems is available, so the reason
 may be
 >> > limited to users using Mac OS. And then, I using the ping the ipv6
 >> address
 >> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as
 the host
 >> > name. based on this default behavior of the operating system, I
 >> recommend
 >> > using macos with ipv4 and others with ipv4/ipv6.
 >> >
 >> > On 2024/01/18 06:54:14 Istvan Toth wrote:
 >> > > I think we should look into this a bit deeper.
 >> > > Why does this fail with IPv6 ?
 >> > >
 >> > > Are we explicitly binding TSO to the IPv4 address only ?
 >> > > Do we only do that in the tests, or does that happen in
 production ?
 >> > > Is it even possible to use Omid with IPv6 ?
 >> > >
 >> > > I suspect that fixing
 https://issues.apache.org/jira/browse/OMID-249
 >> > would
 >> > > improve on this.
 >> > >
 >> > > Generally, I think that having a single IP per TSO can be a
 problem
 >> in a
 >> > > mixed IPv4 / IPv6 network, and with some network configurations.
 >> > > Maybe we should store the name of the active server instead of
 >> addresses
 >> > in
 >> > > ZK, and kick this problem down to the network/DNS administrator.
 >> > >
 >> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK
 with
 >> Cong's
 >> > > proposed solution, but if it breaks pre-existing Ipv6 support,
 then we
 >> > > should
 >> > > fix this properly for 1.1.1.
 >> > >
 >> > > We should create a follow-up ticket to make sure the Omid is IPv6
 >> clean
 >> > for
 >> > > the next release at the very least.
 >> > >
 >> > > Istvan
 >> > >
 >> > >
 >> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo 
 wrote:
 >> > >
 >> > > > Hi Richard,
 >> > > >   Actually, I had the same issue before, but it was resolved
 >> locally:
 >> > if
 >> > > > the NIC was assigned both ipv4 and ipv6 addresses, use ipv4
 first.
 >> If
 >> > we
 >> > > > all think this is acceptable, then I will create a PR
 submission.
 >> > > >
 >> > > > On 2024/01/17 14:30:50 Richárd Antal wrote:
 >> > > > > -1
 >> > > > >
 >> > > > > Signature: OK
 >> > > > > Checksum: OK
 >> > > > > mvn clean apache-rat:check: OK
 >> > > > >
 >> > > > > Changes and release notes: OK
 >> > > > >
 >> > > > > Built from source (1.8.0_241): OK
 >> > > > >   - mvn clean install  -DskipTests
 >> > > > > Ran tests on my MacBook locally:
 >> > > > >   - mvn clean verify : Failed
 >

Re: [CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[Istvan Toth]

We have worked on this with Richard.

I have made some changes to force the test to use IPv6 on my Linux machine,
and it still works fine.
It is some funky mac networking issue:
https://medium.com/@quelgar/java-sockets-broken-for-ipv6-on-mac-5aae72f06b21

Please review https://issues.apache.org/jira/browse/OMID-249, which fixes
the issue (at least on Richard's particular mac)

best regards
Istvan

On Fri, Jan 19, 2024 at 8:07 AM Istvan Toth  wrote:

> Someone who has access to a Mac please try to track down which change
> broke the tests.
>
>
> On Fri, Jan 19, 2024 at 8:05 AM Istvan Toth  wrote:
>
>> PR for JDK17 support is up:
>> https://issues.apache.org/jira/browse/OMID-272
>>
>> On Thu, Jan 18, 2024 at 6:39 PM rajeshb...@apache.org <
>> chrajeshbab...@gmail.com> wrote:
>>
>>> The vote has been canceled to address issues with JDK and Ipv6 interface
>>> in
>>> MacOS
>>>
>>> Thanks,
>>> Rajeshbabu.
>>>
>>> On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
>>> chrajeshbab...@gmail.com> wrote:
>>>
>>> > Sure, let me cancel the RC and create another RC after fixes with the
>>> > network interface and JDK 17.
>>> >
>>> > Thanks, Richard and Istvan for finding issues, and Cong Luo for your
>>> > support.
>>> >
>>> >
>>> >
>>> >
>>> >
>>> > On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth >> >
>>> > wrote:
>>> >
>>> >> I have found another problem, Omid doesn't work with JDK 17.
>>> >> Not necessarily a blocker, but if we do another RC, it would be nice
>>> to
>>> >> include it.
>>> >>
>>> >> I'm also working on adding HBase 3 support, which requires a small API
>>> >> change (that also affects Phoenix)
>>> >> Again, not a blocker, but we'll have to do another release for 5.2
>>> sooner
>>> >> than later for HBase 3 support.
>>> >>
>>> >>
>>> >> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
>>> >>
>>> >> > Hi Istvan,
>>> >> > Using ipv6 on linux operating systems is available, so the reason
>>> may be
>>> >> > limited to users using Mac OS. And then, I using the ping the ipv6
>>> >> address
>>> >> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as the
>>> host
>>> >> > name. based on this default behavior of the operating system, I
>>> >> recommend
>>> >> > using macos with ipv4 and others with ipv4/ipv6.
>>> >> >
>>> >> > On 2024/01/18 06:54:14 Istvan Toth wrote:
>>> >> > > I think we should look into this a bit deeper.
>>> >> > > Why does this fail with IPv6 ?
>>> >> > >
>>> >> > > Are we explicitly binding TSO to the IPv4 address only ?
>>> >> > > Do we only do that in the tests, or does that happen in
>>> production ?
>>> >> > > Is it even possible to use Omid with IPv6 ?
>>> >> > >
>>> >> > > I suspect that fixing
>>> https://issues.apache.org/jira/browse/OMID-249
>>> >> > would
>>> >> > > improve on this.
>>> >> > >
>>> >> > > Generally, I think that having a single IP per TSO can be a
>>> problem
>>> >> in a
>>> >> > > mixed IPv4 / IPv6 network, and with some network configurations.
>>> >> > > Maybe we should store the name of the active server instead of
>>> >> addresses
>>> >> > in
>>> >> > > ZK, and kick this problem down to the network/DNS administrator.
>>> >> > >
>>> >> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK with
>>> >> Cong's
>>> >> > > proposed solution, but if it breaks pre-existing Ipv6 support,
>>> then we
>>> >> > > should
>>> >> > > fix this properly for 1.1.1.
>>> >> > >
>>> >> > > We should create a follow-up ticket to make sure the Omid is IPv6
>>> >> clean
>>> >> > for
>>> >> > > the next release at the very least.
>>> >> > >
>>> >> > > Istvan
>>> >> > >
>>> >> > >
>>> >> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo  wrote:
>>> >> > >
>>> >> > > > Hi Richard,
>>> >> > > >   Actually, I had the same issue before, but it was resolved
>>> >> locally:
>>> >> > if
>>> >> > > > the NIC was assigned both ipv4 and ipv6 addresses, use ipv4
>>> first.
>>> >> If
>>> >> > we
>>> >> > > > all think this is acceptable, then I will create a PR
>>> submission.
>>> >> > > >
>>> >> > > > On 2024/01/17 14:30:50 Richárd Antal wrote:
>>> >> > > > > -1
>>> >> > > > >
>>> >> > > > > Signature: OK
>>> >> > > > > Checksum: OK
>>> >> > > > > mvn clean apache-rat:check: OK
>>> >> > > > >
>>> >> > > > > Changes and release notes: OK
>>> >> > > > >
>>> >> > > > > Built from source (1.8.0_241): OK
>>> >> > > > >   - mvn clean install  -DskipTests
>>> >> > > > > Ran tests on my MacBook locally:
>>> >> > > > >   - mvn clean verify : Failed
>>> >> > > > > TestTSOClientConnectionToTSO had failing tests:
>>> >> > > > > testSuccessfulConnectionToTSOThroughZK and
>>> >> > > > >
>>> testSuccessOfTSOClientReconnectionsToARestartedTSOWithZKPublishing
>>> >> > > > > It might be because it uses IPv6
>>> >> > > > > Some logs
>>> >> > > > >
>>> >> > > > > 2024-01-17T15:17:45,858 INFO
>>> >> > > > > >  [TestNGInvoker-testSuccessfulConnectionToTSOThroughZK()]
>>> >> > > > client.TSOClient:
>>> >> > > > > > * Current TSO host:port found in ZK:
>>> >> > > > > > [fe80:0:0:0:aede:48ff:fe00:

Re: [CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[Istvan Toth]

Someone who has access to a Mac please try to track down which change broke
the tests.


On Fri, Jan 19, 2024 at 8:05 AM Istvan Toth  wrote:

> PR for JDK17 support is up:
> https://issues.apache.org/jira/browse/OMID-272
>
> On Thu, Jan 18, 2024 at 6:39 PM rajeshb...@apache.org <
> chrajeshbab...@gmail.com> wrote:
>
>> The vote has been canceled to address issues with JDK and Ipv6 interface
>> in
>> MacOS
>>
>> Thanks,
>> Rajeshbabu.
>>
>> On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
>> chrajeshbab...@gmail.com> wrote:
>>
>> > Sure, let me cancel the RC and create another RC after fixes with the
>> > network interface and JDK 17.
>> >
>> > Thanks, Richard and Istvan for finding issues, and Cong Luo for your
>> > support.
>> >
>> >
>> >
>> >
>> >
>> > On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth > >
>> > wrote:
>> >
>> >> I have found another problem, Omid doesn't work with JDK 17.
>> >> Not necessarily a blocker, but if we do another RC, it would be nice to
>> >> include it.
>> >>
>> >> I'm also working on adding HBase 3 support, which requires a small API
>> >> change (that also affects Phoenix)
>> >> Again, not a blocker, but we'll have to do another release for 5.2
>> sooner
>> >> than later for HBase 3 support.
>> >>
>> >>
>> >> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
>> >>
>> >> > Hi Istvan,
>> >> > Using ipv6 on linux operating systems is available, so the reason
>> may be
>> >> > limited to users using Mac OS. And then, I using the ping the ipv6
>> >> address
>> >> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as the
>> host
>> >> > name. based on this default behavior of the operating system, I
>> >> recommend
>> >> > using macos with ipv4 and others with ipv4/ipv6.
>> >> >
>> >> > On 2024/01/18 06:54:14 Istvan Toth wrote:
>> >> > > I think we should look into this a bit deeper.
>> >> > > Why does this fail with IPv6 ?
>> >> > >
>> >> > > Are we explicitly binding TSO to the IPv4 address only ?
>> >> > > Do we only do that in the tests, or does that happen in production
>> ?
>> >> > > Is it even possible to use Omid with IPv6 ?
>> >> > >
>> >> > > I suspect that fixing
>> https://issues.apache.org/jira/browse/OMID-249
>> >> > would
>> >> > > improve on this.
>> >> > >
>> >> > > Generally, I think that having a single IP per TSO can be a problem
>> >> in a
>> >> > > mixed IPv4 / IPv6 network, and with some network configurations.
>> >> > > Maybe we should store the name of the active server instead of
>> >> addresses
>> >> > in
>> >> > > ZK, and kick this problem down to the network/DNS administrator.
>> >> > >
>> >> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK with
>> >> Cong's
>> >> > > proposed solution, but if it breaks pre-existing Ipv6 support,
>> then we
>> >> > > should
>> >> > > fix this properly for 1.1.1.
>> >> > >
>> >> > > We should create a follow-up ticket to make sure the Omid is IPv6
>> >> clean
>> >> > for
>> >> > > the next release at the very least.
>> >> > >
>> >> > > Istvan
>> >> > >
>> >> > >
>> >> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo  wrote:
>> >> > >
>> >> > > > Hi Richard,
>> >> > > >   Actually, I had the same issue before, but it was resolved
>> >> locally:
>> >> > if
>> >> > > > the NIC was assigned both ipv4 and ipv6 addresses, use ipv4
>> first.
>> >> If
>> >> > we
>> >> > > > all think this is acceptable, then I will create a PR submission.
>> >> > > >
>> >> > > > On 2024/01/17 14:30:50 Richárd Antal wrote:
>> >> > > > > -1
>> >> > > > >
>> >> > > > > Signature: OK
>> >> > > > > Checksum: OK
>> >> > > > > mvn clean apache-rat:check: OK
>> >> > > > >
>> >> > > > > Changes and release notes: OK
>> >> > > > >
>> >> > > > > Built from source (1.8.0_241): OK
>> >> > > > >   - mvn clean install  -DskipTests
>> >> > > > > Ran tests on my MacBook locally:
>> >> > > > >   - mvn clean verify : Failed
>> >> > > > > TestTSOClientConnectionToTSO had failing tests:
>> >> > > > > testSuccessfulConnectionToTSOThroughZK and
>> >> > > > >
>> testSuccessOfTSOClientReconnectionsToARestartedTSOWithZKPublishing
>> >> > > > > It might be because it uses IPv6
>> >> > > > > Some logs
>> >> > > > >
>> >> > > > > 2024-01-17T15:17:45,858 INFO
>> >> > > > > >  [TestNGInvoker-testSuccessfulConnectionToTSOThroughZK()]
>> >> > > > client.TSOClient:
>> >> > > > > > * Current TSO host:port found in ZK:
>> >> > > > > > [fe80:0:0:0:aede:48ff:fe00:1122%en5]:52934 Epoch 0
>> >> > > > > > 2024-01-17T15:17:45,859 INFO  [tsofsm-0] client.TSOClient:
>> >> Trying
>> >> > to
>> >> > > > > > connect to TSO [/fe80:0:0:0:aede:48ff:fe00:1122%en5:52934]
>> >> > > > > > 2024-01-17T15:17:45,863 ERROR [tsoclient-worker-0]
>> >> > client.TSOClient:
>> >> > > > > > Failed connection attempt to TSO
>> >> > > > > > [/fe80:0:0:0:aede:48ff:fe00:1122%en5:52934] failed. Channel
>> [id:
>> >> > > > 0x9d8c0b44]
>> >> > > > >
>> >> > > > >
>> >> > > > > I’ve also tested the TestTSOClientConnectionToTSO on the 1.1.0
>> >> > release
>> >> > > > > locally and that was

Re: [CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[Istvan Toth]

PR for JDK17 support is up:
https://issues.apache.org/jira/browse/OMID-272

On Thu, Jan 18, 2024 at 6:39 PM rajeshb...@apache.org <
chrajeshbab...@gmail.com> wrote:

> The vote has been canceled to address issues with JDK and Ipv6 interface in
> MacOS
>
> Thanks,
> Rajeshbabu.
>
> On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
> chrajeshbab...@gmail.com> wrote:
>
> > Sure, let me cancel the RC and create another RC after fixes with the
> > network interface and JDK 17.
> >
> > Thanks, Richard and Istvan for finding issues, and Cong Luo for your
> > support.
> >
> >
> >
> >
> >
> > On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth 
> > wrote:
> >
> >> I have found another problem, Omid doesn't work with JDK 17.
> >> Not necessarily a blocker, but if we do another RC, it would be nice to
> >> include it.
> >>
> >> I'm also working on adding HBase 3 support, which requires a small API
> >> change (that also affects Phoenix)
> >> Again, not a blocker, but we'll have to do another release for 5.2
> sooner
> >> than later for HBase 3 support.
> >>
> >>
> >> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
> >>
> >> > Hi Istvan,
> >> > Using ipv6 on linux operating systems is available, so the reason may
> be
> >> > limited to users using Mac OS. And then, I using the ping the ipv6
> >> address
> >> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as the
> host
> >> > name. based on this default behavior of the operating system, I
> >> recommend
> >> > using macos with ipv4 and others with ipv4/ipv6.
> >> >
> >> > On 2024/01/18 06:54:14 Istvan Toth wrote:
> >> > > I think we should look into this a bit deeper.
> >> > > Why does this fail with IPv6 ?
> >> > >
> >> > > Are we explicitly binding TSO to the IPv4 address only ?
> >> > > Do we only do that in the tests, or does that happen in production ?
> >> > > Is it even possible to use Omid with IPv6 ?
> >> > >
> >> > > I suspect that fixing
> https://issues.apache.org/jira/browse/OMID-249
> >> > would
> >> > > improve on this.
> >> > >
> >> > > Generally, I think that having a single IP per TSO can be a problem
> >> in a
> >> > > mixed IPv4 / IPv6 network, and with some network configurations.
> >> > > Maybe we should store the name of the active server instead of
> >> addresses
> >> > in
> >> > > ZK, and kick this problem down to the network/DNS administrator.
> >> > >
> >> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK with
> >> Cong's
> >> > > proposed solution, but if it breaks pre-existing Ipv6 support, then
> we
> >> > > should
> >> > > fix this properly for 1.1.1.
> >> > >
> >> > > We should create a follow-up ticket to make sure the Omid is IPv6
> >> clean
> >> > for
> >> > > the next release at the very least.
> >> > >
> >> > > Istvan
> >> > >
> >> > >
> >> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo  wrote:
> >> > >
> >> > > > Hi Richard,
> >> > > >   Actually, I had the same issue before, but it was resolved
> >> locally:
> >> > if
> >> > > > the NIC was assigned both ipv4 and ipv6 addresses, use ipv4 first.
> >> If
> >> > we
> >> > > > all think this is acceptable, then I will create a PR submission.
> >> > > >
> >> > > > On 2024/01/17 14:30:50 Richárd Antal wrote:
> >> > > > > -1
> >> > > > >
> >> > > > > Signature: OK
> >> > > > > Checksum: OK
> >> > > > > mvn clean apache-rat:check: OK
> >> > > > >
> >> > > > > Changes and release notes: OK
> >> > > > >
> >> > > > > Built from source (1.8.0_241): OK
> >> > > > >   - mvn clean install  -DskipTests
> >> > > > > Ran tests on my MacBook locally:
> >> > > > >   - mvn clean verify : Failed
> >> > > > > TestTSOClientConnectionToTSO had failing tests:
> >> > > > > testSuccessfulConnectionToTSOThroughZK and
> >> > > > >
> testSuccessOfTSOClientReconnectionsToARestartedTSOWithZKPublishing
> >> > > > > It might be because it uses IPv6
> >> > > > > Some logs
> >> > > > >
> >> > > > > 2024-01-17T15:17:45,858 INFO
> >> > > > > >  [TestNGInvoker-testSuccessfulConnectionToTSOThroughZK()]
> >> > > > client.TSOClient:
> >> > > > > > * Current TSO host:port found in ZK:
> >> > > > > > [fe80:0:0:0:aede:48ff:fe00:1122%en5]:52934 Epoch 0
> >> > > > > > 2024-01-17T15:17:45,859 INFO  [tsofsm-0] client.TSOClient:
> >> Trying
> >> > to
> >> > > > > > connect to TSO [/fe80:0:0:0:aede:48ff:fe00:1122%en5:52934]
> >> > > > > > 2024-01-17T15:17:45,863 ERROR [tsoclient-worker-0]
> >> > client.TSOClient:
> >> > > > > > Failed connection attempt to TSO
> >> > > > > > [/fe80:0:0:0:aede:48ff:fe00:1122%en5:52934] failed. Channel
> [id:
> >> > > > 0x9d8c0b44]
> >> > > > >
> >> > > > >
> >> > > > > I’ve also tested the TestTSOClientConnectionToTSO on the 1.1.0
> >> > release
> >> > > > > locally and that was successful (It used IPv4).
> >> > > > > I think this is a regression on Mac and it would be nice to fix
> >> > before
> >> > > > the
> >> > > > > release.
> >> > > > >
> >> > > > > Richard
> >> > > > >
> >> > > > > Cong Luo  ezt írta (időpont: 2024. jan. 11.,
> Cs,
> >> > 3:52):
> >> > > > >
> >

[CANCEL][VOTE] Release of Apache Phoenix Omid 1.1.1 RC0

[rajeshb...@apache.org]

The vote has been canceled to address issues with JDK and Ipv6 interface in
MacOS

Thanks,
Rajeshbabu.

On Thu, Jan 18, 2024 at 11:05 PM rajeshb...@apache.org <
chrajeshbab...@gmail.com> wrote:

> Sure, let me cancel the RC and create another RC after fixes with the
> network interface and JDK 17.
>
> Thanks, Richard and Istvan for finding issues, and Cong Luo for your
> support.
>
>
>
>
>
> On Thu, Jan 18, 2024 at 9:52 PM Istvan Toth 
> wrote:
>
>> I have found another problem, Omid doesn't work with JDK 17.
>> Not necessarily a blocker, but if we do another RC, it would be nice to
>> include it.
>>
>> I'm also working on adding HBase 3 support, which requires a small API
>> change (that also affects Phoenix)
>> Again, not a blocker, but we'll have to do another release for 5.2 sooner
>> than later for HBase 3 support.
>>
>>
>> On Thu, Jan 18, 2024 at 9:11 AM Cong Luo  wrote:
>>
>> > Hi Istvan,
>> > Using ipv6 on linux operating systems is available, so the reason may be
>> > limited to users using Mac OS. And then, I using the ping the ipv6
>> address
>> > on Mac OS, got the 'Unknown host', obviously macos uses ipv6 as the host
>> > name. based on this default behavior of the operating system, I
>> recommend
>> > using macos with ipv4 and others with ipv4/ipv6.
>> >
>> > On 2024/01/18 06:54:14 Istvan Toth wrote:
>> > > I think we should look into this a bit deeper.
>> > > Why does this fail with IPv6 ?
>> > >
>> > > Are we explicitly binding TSO to the IPv4 address only ?
>> > > Do we only do that in the tests, or does that happen in production ?
>> > > Is it even possible to use Omid with IPv6 ?
>> > >
>> > > I suspect that fixing https://issues.apache.org/jira/browse/OMID-249
>> > would
>> > > improve on this.
>> > >
>> > > Generally, I think that having a single IP per TSO can be a problem
>> in a
>> > > mixed IPv4 / IPv6 network, and with some network configurations.
>> > > Maybe we should store the name of the active server instead of
>> addresses
>> > in
>> > > ZK, and kick this problem down to the network/DNS administrator.
>> > >
>> > > If Omid is known to not work with IPV6 in 1.1.0 , then I'm OK with
>> Cong's
>> > > proposed solution, but if it breaks pre-existing Ipv6 support, then we
>> > > should
>> > > fix this properly for 1.1.1.
>> > >
>> > > We should create a follow-up ticket to make sure the Omid is IPv6
>> clean
>> > for
>> > > the next release at the very least.
>> > >
>> > > Istvan
>> > >
>> > >
>> > > On Thu, Jan 18, 2024 at 6:56 AM Cong Luo  wrote:
>> > >
>> > > > Hi Richard,
>> > > >   Actually, I had the same issue before, but it was resolved
>> locally:
>> > if
>> > > > the NIC was assigned both ipv4 and ipv6 addresses, use ipv4 first.
>> If
>> > we
>> > > > all think this is acceptable, then I will create a PR submission.
>> > > >
>> > > > On 2024/01/17 14:30:50 Richárd Antal wrote:
>> > > > > -1
>> > > > >
>> > > > > Signature: OK
>> > > > > Checksum: OK
>> > > > > mvn clean apache-rat:check: OK
>> > > > >
>> > > > > Changes and release notes: OK
>> > > > >
>> > > > > Built from source (1.8.0_241): OK
>> > > > >   - mvn clean install  -DskipTests
>> > > > > Ran tests on my MacBook locally:
>> > > > >   - mvn clean verify : Failed
>> > > > > TestTSOClientConnectionToTSO had failing tests:
>> > > > > testSuccessfulConnectionToTSOThroughZK and
>> > > > > testSuccessOfTSOClientReconnectionsToARestartedTSOWithZKPublishing
>> > > > > It might be because it uses IPv6
>> > > > > Some logs
>> > > > >
>> > > > > 2024-01-17T15:17:45,858 INFO
>> > > > > >  [TestNGInvoker-testSuccessfulConnectionToTSOThroughZK()]
>> > > > client.TSOClient:
>> > > > > > * Current TSO host:port found in ZK:
>> > > > > > [fe80:0:0:0:aede:48ff:fe00:1122%en5]:52934 Epoch 0
>> > > > > > 2024-01-17T15:17:45,859 INFO  [tsofsm-0] client.TSOClient:
>> Trying
>> > to
>> > > > > > connect to TSO [/fe80:0:0:0:aede:48ff:fe00:1122%en5:52934]
>> > > > > > 2024-01-17T15:17:45,863 ERROR [tsoclient-worker-0]
>> > client.TSOClient:
>> > > > > > Failed connection attempt to TSO
>> > > > > > [/fe80:0:0:0:aede:48ff:fe00:1122%en5:52934] failed. Channel [id:
>> > > > 0x9d8c0b44]
>> > > > >
>> > > > >
>> > > > > I’ve also tested the TestTSOClientConnectionToTSO on the 1.1.0
>> > release
>> > > > > locally and that was successful (It used IPv4).
>> > > > > I think this is a regression on Mac and it would be nice to fix
>> > before
>> > > > the
>> > > > > release.
>> > > > >
>> > > > > Richard
>> > > > >
>> > > > > Cong Luo  ezt írta (időpont: 2024. jan. 11., Cs,
>> > 3:52):
>> > > > >
>> > > > > > +1
>> > > > > >
>> > > > > > I did a basic test last month in the test environment (fully
>> > > > distributed)
>> > > > > > based on phoenix 5.1.3 + omid 1.1.1 and did not find a blocking
>> > issue.
>> > > > > >
>> > > > > > On 2024/01/09 12:53:28 "rajeshb...@apache.org" wrote:
>> > > > > > > Please vote on this Apache phoenix omid release candidate,
>> > > > > > > phoenix-omid-1.1.1RC0
>> > > > > > >
>> > > > > > > The VOTE will remai