You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@geode.apache.org by Bruce Schuchardt <bs...@pivotal.io> on 2017/06/08 21:23:43 UTC
release 1.2
I'd like to include the fix for GEODE-3052 in the 1.2 release. It's
under review now and has passed new tests and is in precheckin testing.
https://issues.apache.org/jira/browse/GEODE-3052
"Restarting 2 locators within 1s of each other causes potential locator
split brain"
Re: release 1.2
Posted by Jared Stewart <js...@pivotal.io>.
Sounds good. I believe GEODE-3054 is already fixed on develop and release/1.2.0.
- Jared
> On Jun 10, 2017, at 10:15 AM, Anthony Baker <ab...@pivotal.io> wrote:
>
> Status update:
>
> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
>
> Anthony
>
Re: release 1.2
Posted by Dave Barnes <db...@pivotal.io>.
I have staged a v1.2 doc update for the Apache Geode website, using RC 1
versions of the User Guide and the Javadocs.
Later release candidates are unlikely to require doc changes, other than
the Wiki-based relnotes, so I'm nearly certain this will be the final
version.
When we merge or rename this branch to 'asf-site', Apache will publish it.
Because I may be on vacation when the release goes live, I checked it into
the geode-site repo as a feature branch, feature/Geode-1.2-webpages.
-Dave
PS: Apologies for the awkward name, but I don't know exactly which words
trigger Apache's deus-ex-machina, so thought I'd avoid "asf" and "site".
On Sat, Jun 17, 2017 at 11:43 AM, Anthony Baker <ab...@pivotal.io> wrote:
>
> > On Jun 14, 2017, at 4:59 PM, Anthony Baker <ab...@pivotal.io> wrote:
> >
> > Agreed. Please merge to release/1.2.0 when ready.
> >
> > Anthony
> >
>
> Hi everyone,
>
> Please contribute to the ‘crowd-sourced’ releases notes at:
> https://cwiki.apache.org/confluence/display/GEODE/Release+Notes <
> https://cwiki.apache.org/confluence/display/GEODE/Release+Notes>
>
> If you have contributed something you want to be called out, please add a
> note.
>
> Thanks,
> Anthony
>
>
>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
> On Jun 14, 2017, at 4:59 PM, Anthony Baker <ab...@pivotal.io> wrote:
>
> Agreed. Please merge to release/1.2.0 when ready.
>
> Anthony
>
Hi everyone,
Please contribute to the ‘crowd-sourced’ releases notes at:
https://cwiki.apache.org/confluence/display/GEODE/Release+Notes <https://cwiki.apache.org/confluence/display/GEODE/Release+Notes>
If you have contributed something you want to be called out, please add a note.
Thanks,
Anthony
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
Agreed. Please merge to release/1.2.0 when ready.
Anthony
> On Jun 14, 2017, at 1:46 PM, Lynn Hughes-Godfrey <lh...@pivotal.io> wrote:
>
> We would like to include the fix for (GEODE-3072) Events do not get removed
> from the client queue when 1.0 clients connect to 1.2 servers
>
> On Wed, Jun 14, 2017 at 12:25 PM, Fred Krone <fk...@pivotal.io> wrote:
>
>> Reviewing our JTA transaction manager implementation during a recent issue
>> we thought it would be useful to log message for users when our
>> implementation is being used (which is by default but potentially unknown).
>>
>>
>> It's small but not critical -- we thought we might get it in if the 1.2
>> release lingered. Eric was running pre-check-in on it.
>>
>> On Wed, Jun 14, 2017 at 12:17 PM, Anthony Baker <ab...@pivotal.io> wrote:
>>
>>>
>>>> On Jun 14, 2017, at 12:03 PM, Eric Shu <es...@pivotal.io> wrote:
>>>>
>>>> I'd like to include GEODE-2301 in the release 1.2.0 (deprecate GEODE
>>>> implementation of JTA transaction manager)
>>>>
>>>> -Eric
>>>
>>> Thanks Eric. Can you make a case for why this change needs to go into
>>> 1.2.0? We’re pretty locked down on 1.2.0 changes and unless it’s a
>>> critical issue I’d prefer to not introduce more changes.
>>>
>>> Has the deprecation of this feature been discussed?
>>>
>>> Anthony
>>>
>>>
>>
Re: release 1.2
Posted by Lynn Hughes-Godfrey <lh...@pivotal.io>.
We would like to include the fix for (GEODE-3072) Events do not get removed
from the client queue when 1.0 clients connect to 1.2 servers
On Wed, Jun 14, 2017 at 12:25 PM, Fred Krone <fk...@pivotal.io> wrote:
> Reviewing our JTA transaction manager implementation during a recent issue
> we thought it would be useful to log message for users when our
> implementation is being used (which is by default but potentially unknown).
>
>
> It's small but not critical -- we thought we might get it in if the 1.2
> release lingered. Eric was running pre-check-in on it.
>
> On Wed, Jun 14, 2017 at 12:17 PM, Anthony Baker <ab...@pivotal.io> wrote:
>
> >
> > > On Jun 14, 2017, at 12:03 PM, Eric Shu <es...@pivotal.io> wrote:
> > >
> > > I'd like to include GEODE-2301 in the release 1.2.0 (deprecate GEODE
> > > implementation of JTA transaction manager)
> > >
> > > -Eric
> >
> > Thanks Eric. Can you make a case for why this change needs to go into
> > 1.2.0? We’re pretty locked down on 1.2.0 changes and unless it’s a
> > critical issue I’d prefer to not introduce more changes.
> >
> > Has the deprecation of this feature been discussed?
> >
> > Anthony
> >
> >
>
Re: release 1.2
Posted by Fred Krone <fk...@pivotal.io>.
Reviewing our JTA transaction manager implementation during a recent issue
we thought it would be useful to log message for users when our
implementation is being used (which is by default but potentially unknown).
It's small but not critical -- we thought we might get it in if the 1.2
release lingered. Eric was running pre-check-in on it.
On Wed, Jun 14, 2017 at 12:17 PM, Anthony Baker <ab...@pivotal.io> wrote:
>
> > On Jun 14, 2017, at 12:03 PM, Eric Shu <es...@pivotal.io> wrote:
> >
> > I'd like to include GEODE-2301 in the release 1.2.0 (deprecate GEODE
> > implementation of JTA transaction manager)
> >
> > -Eric
>
> Thanks Eric. Can you make a case for why this change needs to go into
> 1.2.0? We’re pretty locked down on 1.2.0 changes and unless it’s a
> critical issue I’d prefer to not introduce more changes.
>
> Has the deprecation of this feature been discussed?
>
> Anthony
>
>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
> On Jun 14, 2017, at 12:03 PM, Eric Shu <es...@pivotal.io> wrote:
>
> I'd like to include GEODE-2301 in the release 1.2.0 (deprecate GEODE
> implementation of JTA transaction manager)
>
> -Eric
Thanks Eric. Can you make a case for why this change needs to go into 1.2.0? We’re pretty locked down on 1.2.0 changes and unless it’s a critical issue I’d prefer to not introduce more changes.
Has the deprecation of this feature been discussed?
Anthony
Re: release 1.2
Posted by Eric Shu <es...@pivotal.io>.
I'd like to include GEODE-2301 in the release 1.2.0 (deprecate GEODE
implementation of JTA transaction manager)
-Eric
On Wed, Jun 14, 2017 at 11:50 AM, Anthony Baker <ab...@pivotal.io> wrote:
> Status update:
>
> - I think we have all changes needed in the release/1.2.0 branch [1].
> Please shout out if not.
> - The last 2 Geode-release jenkins runs have timed out / hung [2]. Please
> review to see if there are changes needed.
> - I’ve completed the LICENSE review [3].
>
> Anthony
>
> [1] https://github.com/apache/geode/tree/release/1.2.0 <
> https://github.com/apache/geode/tree/release/1.2.0>
> [2] https://builds.apache.org/job/Geode-release/ <
> https://builds.apache.org/job/Geode-release/>
> [3] https://github.com/apache/geode/commit/24192d4fbc8739d0f4dae2b4de5c8f
> d00b82ad32 <https://github.com/apache/geode/commit/
> 24192d4fbc8739d0f4dae2b4de5c8fd00b82ad32>
>
>
> > On Jun 13, 2017, at 9:44 AM, Anthony Baker <ab...@pivotal.io> wrote:
> >
> > Vahram,
> >
> > Thanks for advocating for this issue! While the release it pretty
> locked down for changes, this looks like a low risk change for an important
> issue. I’ll cherry-pick the changes onto release/1.2.0.
> >
> > I also see that a breaking change to the REST API was made in
> ‘c39585395a’. Can someone take a look at the message vs cause naming for
> errors?
> >
> > Anthony
> >
> >> On Jun 12, 2017, at 9:43 AM, Vahram Aharonyan <vaharonyan@vmware.com
> <ma...@vmware.com>> wrote:
> >>
> >> Hi All,
> >>
> >> We really think fix done for GEODE-3023 should be considered for 1.2
> version as well. Especially since similar fix(GEODE-2898) has been already
> made it to 1.2 version. So shipping a partial/incomplete fix in 1.2 version
> for the same kind of problem does not really provide value, and still
> leaves the system exposed with a resiliency problem in case of network
> problems.
> >>
> >> The problem is indeed critical, it is hitting on nearly every
> meaningfully large environment which has reasonable amount of network
> issues. It turns out that this problem is impacting entire Geode cluster,
> whenever there is Geode client with let say imperfect quality of network
> connectivity. To overcome this SSL should be disabled for the cluster,
> which is not the best way to proceed.
> >>
> >> Thanks,
> >> Vahram.
> >>
> >> -----Original Message-----
> >> From: Anthony Baker [mailto:abaker@pivotal.io <mailto:abaker@pivotal.io
> >]
> >> Sent: Monday, June 12, 2017 7:42 PM
> >> To: dev@geode.apache.org <ma...@geode.apache.org>
> >> Subject: Re: release 1.2
> >>
> >> Got it, thanks!
> >>
> >> Anthony
> >>
> >>> On Jun 10, 2017, at 7:15 PM, Udo Kohlmeyer <ukohlmeyer@pivotal.io
> <ma...@pivotal.io>> wrote:
> >>>
> >>> I thought we had moved GEODE-3052 to a later release, given that we
> have a work around for this.
> >>>
> >>> --Udo
> >>>
> >>>
> >>> On 6/10/17 10:15, Anthony Baker wrote:
> >>>> Status update:
> >>>>
> >>>> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a
> clean run of the Geode-release jenkins job. So as soon as we can finish
> off these issues and get another clean run through Jenkins I’ll create a
> release candidate. How does early next week sound?
> >>>>
> >>>> Anthony
> >>>>
> >>>>> On Jun 8, 2017, at 5:38 PM, Anthony Baker <abaker@pivotal.io
> <ma...@pivotal.io>> wrote:
> >>>>>
> >>>>> Sounds good. Please cherry-pick these changes onto release/1.2.0
> when ready.
> >>>>>
> >>>>> Anthony
> >>>>>
> >>>>>
> >>>>>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <jstewart@pivotal.io
> <ma...@pivotal.io>> wrote:
> >>>>>>
> >>>>>> I think we probably ought to include GEODE-3045 <
> https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__issues.apache.org_jira_browse_GEODE-2D3045&d=DwIFaQ&
> c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-
> RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e= <https://urldefense.
> proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_
> browse_GEODE-2D3045&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-
> RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e=> > as well.
> >>>>>>
> >>>>>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <khowe@pivotal.io
> <ma...@pivotal.io>> wrote:
> >>>>>>>
> >>>>>>> I think we should also include the fix for GEODE-3054 in 1.2. This
> is a bug that was introduced in the recent changes to the Gfsh parser, and
> affects Windows pathnames specified in options to gfsh commands.
> >>>>>>>
> >>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.
> apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=
> jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.
> proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_
> browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=
> jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e=> <https://urldefense.
> proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_
> browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=
> jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.
> proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_
> browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=
> jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e=> >
> >>>>>>>
> >>>>>>> The fix is currently in testing.
> >>>>>>>
> >>>>>>>
> >>>>>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <
> bschuchardt@pivotal.io <ma...@pivotal.io>> wrote:
> >>>>>>>>
> >>>>>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release.
> It's under review now and has passed new tests and is in precheckin testing.
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.
> apache.org_jira_browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_
> X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e= <https://urldefense.
> proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_
> browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=
> wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=
> 3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_
> X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e=>
> >>>>>>>>
> >>>>>>>> "Restarting 2 locators within 1s of each other causes potential
> locator split brain"
> >
>
>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
Status update:
- I think we have all changes needed in the release/1.2.0 branch [1]. Please shout out if not.
- The last 2 Geode-release jenkins runs have timed out / hung [2]. Please review to see if there are changes needed.
- I’ve completed the LICENSE review [3].
Anthony
[1] https://github.com/apache/geode/tree/release/1.2.0 <https://github.com/apache/geode/tree/release/1.2.0>
[2] https://builds.apache.org/job/Geode-release/ <https://builds.apache.org/job/Geode-release/>
[3] https://github.com/apache/geode/commit/24192d4fbc8739d0f4dae2b4de5c8fd00b82ad32 <https://github.com/apache/geode/commit/24192d4fbc8739d0f4dae2b4de5c8fd00b82ad32>
> On Jun 13, 2017, at 9:44 AM, Anthony Baker <ab...@pivotal.io> wrote:
>
> Vahram,
>
> Thanks for advocating for this issue! While the release it pretty locked down for changes, this looks like a low risk change for an important issue. I’ll cherry-pick the changes onto release/1.2.0.
>
> I also see that a breaking change to the REST API was made in ‘c39585395a’. Can someone take a look at the message vs cause naming for errors?
>
> Anthony
>
>> On Jun 12, 2017, at 9:43 AM, Vahram Aharonyan <vaharonyan@vmware.com <ma...@vmware.com>> wrote:
>>
>> Hi All,
>>
>> We really think fix done for GEODE-3023 should be considered for 1.2 version as well. Especially since similar fix(GEODE-2898) has been already made it to 1.2 version. So shipping a partial/incomplete fix in 1.2 version for the same kind of problem does not really provide value, and still leaves the system exposed with a resiliency problem in case of network problems.
>>
>> The problem is indeed critical, it is hitting on nearly every meaningfully large environment which has reasonable amount of network issues. It turns out that this problem is impacting entire Geode cluster, whenever there is Geode client with let say imperfect quality of network connectivity. To overcome this SSL should be disabled for the cluster, which is not the best way to proceed.
>>
>> Thanks,
>> Vahram.
>>
>> -----Original Message-----
>> From: Anthony Baker [mailto:abaker@pivotal.io <ma...@pivotal.io>]
>> Sent: Monday, June 12, 2017 7:42 PM
>> To: dev@geode.apache.org <ma...@geode.apache.org>
>> Subject: Re: release 1.2
>>
>> Got it, thanks!
>>
>> Anthony
>>
>>> On Jun 10, 2017, at 7:15 PM, Udo Kohlmeyer <ukohlmeyer@pivotal.io <ma...@pivotal.io>> wrote:
>>>
>>> I thought we had moved GEODE-3052 to a later release, given that we have a work around for this.
>>>
>>> --Udo
>>>
>>>
>>> On 6/10/17 10:15, Anthony Baker wrote:
>>>> Status update:
>>>>
>>>> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
>>>>
>>>> Anthony
>>>>
>>>>> On Jun 8, 2017, at 5:38 PM, Anthony Baker <abaker@pivotal.io <ma...@pivotal.io>> wrote:
>>>>>
>>>>> Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
>>>>>
>>>>> Anthony
>>>>>
>>>>>
>>>>>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <jstewart@pivotal.io <ma...@pivotal.io>> wrote:
>>>>>>
>>>>>> I think we probably ought to include GEODE-3045 <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3045&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3045&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e=> > as well.
>>>>>>
>>>>>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <khowe@pivotal.io <ma...@pivotal.io>> wrote:
>>>>>>>
>>>>>>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>>>>>>
>>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e=> <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e=> >
>>>>>>>
>>>>>>> The fix is currently in testing.
>>>>>>>
>>>>>>>
>>>>>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bschuchardt@pivotal.io <ma...@pivotal.io>> wrote:
>>>>>>>>
>>>>>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>>>>>>
>>>>>>>>
>>>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e=>
>>>>>>>>
>>>>>>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
Vahram,
Thanks for advocating for this issue! While the release it pretty locked down for changes, this looks like a low risk change for an important issue. I’ll cherry-pick the changes onto release/1.2.0.
I also see that a breaking change to the REST API was made in ‘c39585395a’. Can someone take a look at the message vs cause naming for errors?
Anthony
> On Jun 12, 2017, at 9:43 AM, Vahram Aharonyan <va...@vmware.com> wrote:
>
> Hi All,
>
> We really think fix done for GEODE-3023 should be considered for 1.2 version as well. Especially since similar fix(GEODE-2898) has been already made it to 1.2 version. So shipping a partial/incomplete fix in 1.2 version for the same kind of problem does not really provide value, and still leaves the system exposed with a resiliency problem in case of network problems.
>
> The problem is indeed critical, it is hitting on nearly every meaningfully large environment which has reasonable amount of network issues. It turns out that this problem is impacting entire Geode cluster, whenever there is Geode client with let say imperfect quality of network connectivity. To overcome this SSL should be disabled for the cluster, which is not the best way to proceed.
>
> Thanks,
> Vahram.
>
> -----Original Message-----
> From: Anthony Baker [mailto:abaker@pivotal.io <ma...@pivotal.io>]
> Sent: Monday, June 12, 2017 7:42 PM
> To: dev@geode.apache.org <ma...@geode.apache.org>
> Subject: Re: release 1.2
>
> Got it, thanks!
>
> Anthony
>
>> On Jun 10, 2017, at 7:15 PM, Udo Kohlmeyer <uk...@pivotal.io> wrote:
>>
>> I thought we had moved GEODE-3052 to a later release, given that we have a work around for this.
>>
>> --Udo
>>
>>
>> On 6/10/17 10:15, Anthony Baker wrote:
>>> Status update:
>>>
>>> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
>>>
>>> Anthony
>>>
>>>> On Jun 8, 2017, at 5:38 PM, Anthony Baker <ab...@pivotal.io> wrote:
>>>>
>>>> Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
>>>>
>>>> Anthony
>>>>
>>>>
>>>>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <js...@pivotal.io> wrote:
>>>>>
>>>>> I think we probably ought to include GEODE-3045 <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3045&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3045&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e=> > as well.
>>>>>
>>>>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <khowe@pivotal.io <ma...@pivotal.io>> wrote:
>>>>>>
>>>>>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>>>>>
>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e=> <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e=> >
>>>>>>
>>>>>> The fix is currently in testing.
>>>>>>
>>>>>>
>>>>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bschuchardt@pivotal.io <ma...@pivotal.io>> wrote:
>>>>>>>
>>>>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>>>>>
>>>>>>>
>>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e=>
>>>>>>>
>>>>>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
RE: release 1.2
Posted by Vahram Aharonyan <va...@vmware.com>.
Hi All,
We really think fix done for GEODE-3023 should be considered for 1.2 version as well. Especially since similar fix(GEODE-2898) has been already made it to 1.2 version. So shipping a partial/incomplete fix in 1.2 version for the same kind of problem does not really provide value, and still leaves the system exposed with a resiliency problem in case of network problems.
The problem is indeed critical, it is hitting on nearly every meaningfully large environment which has reasonable amount of network issues. It turns out that this problem is impacting entire Geode cluster, whenever there is Geode client with let say imperfect quality of network connectivity. To overcome this SSL should be disabled for the cluster, which is not the best way to proceed.
Thanks,
Vahram.
-----Original Message-----
From: Anthony Baker [mailto:abaker@pivotal.io]
Sent: Monday, June 12, 2017 7:42 PM
To: dev@geode.apache.org
Subject: Re: release 1.2
Got it, thanks!
Anthony
> On Jun 10, 2017, at 7:15 PM, Udo Kohlmeyer <uk...@pivotal.io> wrote:
>
> I thought we had moved GEODE-3052 to a later release, given that we have a work around for this.
>
> --Udo
>
>
> On 6/10/17 10:15, Anthony Baker wrote:
>> Status update:
>>
>> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
>>
>> Anthony
>>
>>> On Jun 8, 2017, at 5:38 PM, Anthony Baker <ab...@pivotal.io> wrote:
>>>
>>> Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
>>>
>>> Anthony
>>>
>>>
>>>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <js...@pivotal.io> wrote:
>>>>
>>>> I think we probably ought to include GEODE-3045 <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3045&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=9jqjvxoQ-RJtsGGvTMldMcrooymcWVqjQhb4GoQ1GQ8&e= > as well.
>>>>
>>>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <kh...@pivotal.io> wrote:
>>>>>
>>>>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>>>>
>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3054&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=jSeNlEnjklIGLNnCF8I6cG0t06n2NT5xLBe86ewSmXc&e= >
>>>>>
>>>>> The fix is currently in testing.
>>>>>
>>>>>
>>>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>>>>>>
>>>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>>>>
>>>>>>
>>>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__issues.apache.org_jira_browse_GEODE-2D3052&d=DwIFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=wpTWSXVvcGFCkFEMePbOecdHHTbyiIj9aWq7oqKb0J8&m=3eSDKrOEyEHtI5Oe9-iROOP-AWX9d-a1B-s9UX41yU4&s=T_2Qx57Vhl2-Z_X8qw8OOSwcKYzGBg0IFafSWdbIMPs&e=
>>>>>>
>>>>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>>>>>>
>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
Got it, thanks!
Anthony
> On Jun 10, 2017, at 7:15 PM, Udo Kohlmeyer <uk...@pivotal.io> wrote:
>
> I thought we had moved GEODE-3052 to a later release, given that we have a work around for this.
>
> --Udo
>
>
> On 6/10/17 10:15, Anthony Baker wrote:
>> Status update:
>>
>> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
>>
>> Anthony
>>
>>> On Jun 8, 2017, at 5:38 PM, Anthony Baker <ab...@pivotal.io> wrote:
>>>
>>> Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
>>>
>>> Anthony
>>>
>>>
>>>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <js...@pivotal.io> wrote:
>>>>
>>>> I think we probably ought to include GEODE-3045 <https://issues.apache.org/jira/browse/GEODE-3045> as well.
>>>>
>>>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <kh...@pivotal.io> wrote:
>>>>>
>>>>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>>>>
>>>>> https://issues.apache.org/jira/browse/GEODE-3054 <https://issues.apache.org/jira/browse/GEODE-3054>
>>>>>
>>>>> The fix is currently in testing.
>>>>>
>>>>>
>>>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>>>>>>
>>>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>>>>
>>>>>>
>>>>>> https://issues.apache.org/jira/browse/GEODE-3052
>>>>>>
>>>>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>>>>>>
>
Re: release 1.2
Posted by Udo Kohlmeyer <uk...@pivotal.io>.
I thought we had moved GEODE-3052 to a later release, given that we have
a work around for this.
--Udo
On 6/10/17 10:15, Anthony Baker wrote:
> Status update:
>
> We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
>
> Anthony
>
>> On Jun 8, 2017, at 5:38 PM, Anthony Baker <ab...@pivotal.io> wrote:
>>
>> Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
>>
>> Anthony
>>
>>
>>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <js...@pivotal.io> wrote:
>>>
>>> I think we probably ought to include GEODE-3045 <https://issues.apache.org/jira/browse/GEODE-3045> as well.
>>>
>>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <kh...@pivotal.io> wrote:
>>>>
>>>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>>>
>>>> https://issues.apache.org/jira/browse/GEODE-3054 <https://issues.apache.org/jira/browse/GEODE-3054>
>>>>
>>>> The fix is currently in testing.
>>>>
>>>>
>>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>>>>>
>>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>>>
>>>>>
>>>>> https://issues.apache.org/jira/browse/GEODE-3052
>>>>>
>>>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>>>>>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
Status update:
We have 2 issues left: GEODE-3054 and GEODE-3052. We also had a clean run of the Geode-release jenkins job. So as soon as we can finish off these issues and get another clean run through Jenkins I’ll create a release candidate. How does early next week sound?
Anthony
> On Jun 8, 2017, at 5:38 PM, Anthony Baker <ab...@pivotal.io> wrote:
>
> Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
>
> Anthony
>
>
>> On Jun 8, 2017, at 3:15 PM, Jared Stewart <js...@pivotal.io> wrote:
>>
>> I think we probably ought to include GEODE-3045 <https://issues.apache.org/jira/browse/GEODE-3045> as well.
>>
>>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <kh...@pivotal.io> wrote:
>>>
>>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>>
>>> https://issues.apache.org/jira/browse/GEODE-3054 <https://issues.apache.org/jira/browse/GEODE-3054>
>>>
>>> The fix is currently in testing.
>>>
>>>
>>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>>>>
>>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>>
>>>>
>>>> https://issues.apache.org/jira/browse/GEODE-3052
>>>>
>>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>>>>
>>>
>>
>
Re: release 1.2
Posted by Anthony Baker <ab...@pivotal.io>.
Sounds good. Please cherry-pick these changes onto release/1.2.0 when ready.
Anthony
> On Jun 8, 2017, at 3:15 PM, Jared Stewart <js...@pivotal.io> wrote:
>
> I think we probably ought to include GEODE-3045 <https://issues.apache.org/jira/browse/GEODE-3045> as well.
>
>> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <kh...@pivotal.io> wrote:
>>
>> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>>
>> https://issues.apache.org/jira/browse/GEODE-3054 <https://issues.apache.org/jira/browse/GEODE-3054>
>>
>> The fix is currently in testing.
>>
>>
>>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>>>
>>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>>
>>>
>>> https://issues.apache.org/jira/browse/GEODE-3052
>>>
>>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>>>
>>
>
Re: release 1.2
Posted by Jared Stewart <js...@pivotal.io>.
I think we probably ought to include GEODE-3045 <https://issues.apache.org/jira/browse/GEODE-3045> as well.
> On Jun 8, 2017, at 3:12 PM, Kenneth Howe <kh...@pivotal.io> wrote:
>
> I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
>
> https://issues.apache.org/jira/browse/GEODE-3054 <https://issues.apache.org/jira/browse/GEODE-3054>
>
> The fix is currently in testing.
>
>
>> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>>
>> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>>
>>
>> https://issues.apache.org/jira/browse/GEODE-3052
>>
>> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>>
>
Re: release 1.2
Posted by Kenneth Howe <kh...@pivotal.io>.
I think we should also include the fix for GEODE-3054 in 1.2. This is a bug that was introduced in the recent changes to the Gfsh parser, and affects Windows pathnames specified in options to gfsh commands.
https://issues.apache.org/jira/browse/GEODE-3054 <https://issues.apache.org/jira/browse/GEODE-3054>
The fix is currently in testing.
> On Jun 8, 2017, at 2:23 PM, Bruce Schuchardt <bs...@pivotal.io> wrote:
>
> I'd like to include the fix for GEODE-3052 in the 1.2 release. It's under review now and has passed new tests and is in precheckin testing.
>
>
> https://issues.apache.org/jira/browse/GEODE-3052
>
> "Restarting 2 locators within 1s of each other causes potential locator split brain"
>