Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Dawid Weiss]

bq. because it used to fail sometimes on slow machines

It will very likely fail then on those virtboxes, etc.

D.

On Thu, Mar 15, 2018 at 5:02 AM, Shalin Shekhar Mangar
 wrote:
> Hi Erick,
>
> The test was disabled because it used to fail sometimes on slow machines. I
> haven't beasted it to see if that still happens but it probably does. I only
> fixed it so that it doesn't always fail. So let's see how Jenkins is doing
> and then decide.
>
> On Thu, Mar 15, 2018 at 3:31 AM, Erick Erickson 
> wrote:
>>
>> Shalin:
>>
>> Should we remove (actually, comment out with a date?) the BadApple
>> annotation for doTestIndexFetchOnMasterRestart? And do you think your
>> fixes have any influence on the other BadApple
>> (doTestIndexAndConfigReplication)?
>>
>> There's no problem with un-BadApple-ing test that have been or are
>> being worked on, and we'd get more test coverage that way.
>>
>> Or I can do that on Saturday if you'd prefer, assuming the Jenkins
>> BadApple tests don't show failures.
>>
>>
>>
>> On Wed, Mar 14, 2018 at 1:57 PM, Shalin Shekhar Mangar
>>  wrote:
>> > This is fixed. I committed the fix to master, branch_7x and branch_7_3
>> > branches.
>> >
>> > On Wed, Mar 14, 2018 at 9:44 PM, Alan Woodward 
>> > wrote:
>> >>
>> >> Thanks Shalin!
>> >>
>> >>
>> >> On 14 Mar 2018, at 15:50, Shalin Shekhar Mangar
>> >> 
>> >> wrote:
>> >>
>> >> I'll take a look at it tomorrow morning my time.
>> >>
>> >> On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki
>> >>  wrote:
>> >>>
>> >>> Well … I looked at it briefly but I have no idea what’s going on
>> >>> there. I
>> >>> could dig into it nonetheless, but if there’s someone who already
>> >>> knows the
>> >>> replication handler ins and outs it would probably get fixed sooner...
>> >>>
>> >>>
>> >>> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
>> >>>
>> >>> I’m happy either way, but if it’s a bug can we get it fixed quickly?
>> >>> Can
>> >>> you take ownership of this one Andrzej?
>> >>>
>> >>> On 14 Mar 2018, at 11:24, Andrzej Białecki  wrote:
>> >>>
>> >>> Hi,
>> >>>
>> >>> This test has always been fragile, but recently it’s been failing
>> >>> 100%,
>> >>> most often in ‘doTestIndexFetchOnMasterRestart’.
>> >>>
>> >>> I don’t know the replication handler enough to be able to find the
>> >>> real
>> >>> reason behind these failures, but there are two possibilities that I
>> >>> see:
>> >>>
>> >>> * the test has a bug and needs to be fixed - and if we can’t fix it
>> >>> soon
>> >>> then with 7.3 release imminent we could BadApple it until it’s
>> >>> properly
>> >>> fixed
>> >>>
>> >>> * or actually the replication handler has a bug, which needs to be
>> >>> fixed
>> >>> - in which case I propose to bump up SOLR-12078 to Blocker.
>> >>>
>> >>> I’m open to suggestions.
>> >>>
>> >>> —
>> >>>
>> >>> Andrzej Białecki
>> >>>
>> >>>
>> >>>
>> >>
>> >>
>> >>
>> >> --
>> >> Regards,
>> >> Shalin Shekhar Mangar.
>> >>
>> >>
>> >
>> >
>> >
>> > --
>> > Regards,
>> > Shalin Shekhar Mangar.
>>
>> -
>> To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
>> For additional commands, e-mail: dev-h...@lucene.apache.org
>>
>
>
>
> --
> Regards,
> Shalin Shekhar Mangar.

-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Erick Erickson]

OK. Note that only certain Jenkins jobs run with BadApple=true set.

Mark's beastit stuff (he's got that back up and running) runs with
badapple=true though, so maybe if it survives a few days running there
we can un-annotate it.

How about this. I'll check Marks stuff and, if by Saturday there are
no failures for that test, un-BadApple it on Saturday?

Erick

On Wed, Mar 14, 2018 at 9:02 PM, Shalin Shekhar Mangar
 wrote:
> Hi Erick,
>
> The test was disabled because it used to fail sometimes on slow machines. I
> haven't beasted it to see if that still happens but it probably does. I only
> fixed it so that it doesn't always fail. So let's see how Jenkins is doing
> and then decide.
>
> On Thu, Mar 15, 2018 at 3:31 AM, Erick Erickson 
> wrote:
>>
>> Shalin:
>>
>> Should we remove (actually, comment out with a date?) the BadApple
>> annotation for doTestIndexFetchOnMasterRestart? And do you think your
>> fixes have any influence on the other BadApple
>> (doTestIndexAndConfigReplication)?
>>
>> There's no problem with un-BadApple-ing test that have been or are
>> being worked on, and we'd get more test coverage that way.
>>
>> Or I can do that on Saturday if you'd prefer, assuming the Jenkins
>> BadApple tests don't show failures.
>>
>>
>>
>> On Wed, Mar 14, 2018 at 1:57 PM, Shalin Shekhar Mangar
>>  wrote:
>> > This is fixed. I committed the fix to master, branch_7x and branch_7_3
>> > branches.
>> >
>> > On Wed, Mar 14, 2018 at 9:44 PM, Alan Woodward 
>> > wrote:
>> >>
>> >> Thanks Shalin!
>> >>
>> >>
>> >> On 14 Mar 2018, at 15:50, Shalin Shekhar Mangar
>> >> 
>> >> wrote:
>> >>
>> >> I'll take a look at it tomorrow morning my time.
>> >>
>> >> On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki
>> >>  wrote:
>> >>>
>> >>> Well … I looked at it briefly but I have no idea what’s going on
>> >>> there. I
>> >>> could dig into it nonetheless, but if there’s someone who already
>> >>> knows the
>> >>> replication handler ins and outs it would probably get fixed sooner...
>> >>>
>> >>>
>> >>> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
>> >>>
>> >>> I’m happy either way, but if it’s a bug can we get it fixed quickly?
>> >>> Can
>> >>> you take ownership of this one Andrzej?
>> >>>
>> >>> On 14 Mar 2018, at 11:24, Andrzej Białecki  wrote:
>> >>>
>> >>> Hi,
>> >>>
>> >>> This test has always been fragile, but recently it’s been failing
>> >>> 100%,
>> >>> most often in ‘doTestIndexFetchOnMasterRestart’.
>> >>>
>> >>> I don’t know the replication handler enough to be able to find the
>> >>> real
>> >>> reason behind these failures, but there are two possibilities that I
>> >>> see:
>> >>>
>> >>> * the test has a bug and needs to be fixed - and if we can’t fix it
>> >>> soon
>> >>> then with 7.3 release imminent we could BadApple it until it’s
>> >>> properly
>> >>> fixed
>> >>>
>> >>> * or actually the replication handler has a bug, which needs to be
>> >>> fixed
>> >>> - in which case I propose to bump up SOLR-12078 to Blocker.
>> >>>
>> >>> I’m open to suggestions.
>> >>>
>> >>> —
>> >>>
>> >>> Andrzej Białecki
>> >>>
>> >>>
>> >>>
>> >>
>> >>
>> >>
>> >> --
>> >> Regards,
>> >> Shalin Shekhar Mangar.
>> >>
>> >>
>> >
>> >
>> >
>> > --
>> > Regards,
>> > Shalin Shekhar Mangar.
>>
>> -
>> To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
>> For additional commands, e-mail: dev-h...@lucene.apache.org
>>
>
>
>
> --
> Regards,
> Shalin Shekhar Mangar.

-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Shalin Shekhar Mangar]

Hi Erick,

The test was disabled because it used to fail sometimes on slow machines. I
haven't beasted it to see if that still happens but it probably does. I
only fixed it so that it doesn't always fail. So let's see how Jenkins is
doing and then decide.

On Thu, Mar 15, 2018 at 3:31 AM, Erick Erickson 
wrote:

> Shalin:
>
> Should we remove (actually, comment out with a date?) the BadApple
> annotation for doTestIndexFetchOnMasterRestart? And do you think your
> fixes have any influence on the other BadApple
> (doTestIndexAndConfigReplication)?
>
> There's no problem with un-BadApple-ing test that have been or are
> being worked on, and we'd get more test coverage that way.
>
> Or I can do that on Saturday if you'd prefer, assuming the Jenkins
> BadApple tests don't show failures.
>
>
>
> On Wed, Mar 14, 2018 at 1:57 PM, Shalin Shekhar Mangar
>  wrote:
> > This is fixed. I committed the fix to master, branch_7x and branch_7_3
> > branches.
> >
> > On Wed, Mar 14, 2018 at 9:44 PM, Alan Woodward 
> wrote:
> >>
> >> Thanks Shalin!
> >>
> >>
> >> On 14 Mar 2018, at 15:50, Shalin Shekhar Mangar  >
> >> wrote:
> >>
> >> I'll take a look at it tomorrow morning my time.
> >>
> >> On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki
> >>  wrote:
> >>>
> >>> Well … I looked at it briefly but I have no idea what’s going on
> there. I
> >>> could dig into it nonetheless, but if there’s someone who already
> knows the
> >>> replication handler ins and outs it would probably get fixed sooner...
> >>>
> >>>
> >>> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
> >>>
> >>> I’m happy either way, but if it’s a bug can we get it fixed quickly?
> Can
> >>> you take ownership of this one Andrzej?
> >>>
> >>> On 14 Mar 2018, at 11:24, Andrzej Białecki  wrote:
> >>>
> >>> Hi,
> >>>
> >>> This test has always been fragile, but recently it’s been failing 100%,
> >>> most often in ‘doTestIndexFetchOnMasterRestart’.
> >>>
> >>> I don’t know the replication handler enough to be able to find the real
> >>> reason behind these failures, but there are two possibilities that I
> see:
> >>>
> >>> * the test has a bug and needs to be fixed - and if we can’t fix it
> soon
> >>> then with 7.3 release imminent we could BadApple it until it’s properly
> >>> fixed
> >>>
> >>> * or actually the replication handler has a bug, which needs to be
> fixed
> >>> - in which case I propose to bump up SOLR-12078 to Blocker.
> >>>
> >>> I’m open to suggestions.
> >>>
> >>> —
> >>>
> >>> Andrzej Białecki
> >>>
> >>>
> >>>
> >>
> >>
> >>
> >> --
> >> Regards,
> >> Shalin Shekhar Mangar.
> >>
> >>
> >
> >
> >
> > --
> > Regards,
> > Shalin Shekhar Mangar.
>
> -
> To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
> For additional commands, e-mail: dev-h...@lucene.apache.org
>
>


-- 
Regards,
Shalin Shekhar Mangar.

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Erick Erickson]

Shalin:

Should we remove (actually, comment out with a date?) the BadApple
annotation for doTestIndexFetchOnMasterRestart? And do you think your
fixes have any influence on the other BadApple
(doTestIndexAndConfigReplication)?

There's no problem with un-BadApple-ing test that have been or are
being worked on, and we'd get more test coverage that way.

Or I can do that on Saturday if you'd prefer, assuming the Jenkins
BadApple tests don't show failures.



On Wed, Mar 14, 2018 at 1:57 PM, Shalin Shekhar Mangar
 wrote:
> This is fixed. I committed the fix to master, branch_7x and branch_7_3
> branches.
>
> On Wed, Mar 14, 2018 at 9:44 PM, Alan Woodward  wrote:
>>
>> Thanks Shalin!
>>
>>
>> On 14 Mar 2018, at 15:50, Shalin Shekhar Mangar 
>> wrote:
>>
>> I'll take a look at it tomorrow morning my time.
>>
>> On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki
>>  wrote:
>>>
>>> Well … I looked at it briefly but I have no idea what’s going on there. I
>>> could dig into it nonetheless, but if there’s someone who already knows the
>>> replication handler ins and outs it would probably get fixed sooner...
>>>
>>>
>>> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
>>>
>>> I’m happy either way, but if it’s a bug can we get it fixed quickly?  Can
>>> you take ownership of this one Andrzej?
>>>
>>> On 14 Mar 2018, at 11:24, Andrzej Białecki  wrote:
>>>
>>> Hi,
>>>
>>> This test has always been fragile, but recently it’s been failing 100%,
>>> most often in ‘doTestIndexFetchOnMasterRestart’.
>>>
>>> I don’t know the replication handler enough to be able to find the real
>>> reason behind these failures, but there are two possibilities that I see:
>>>
>>> * the test has a bug and needs to be fixed - and if we can’t fix it soon
>>> then with 7.3 release imminent we could BadApple it until it’s properly
>>> fixed
>>>
>>> * or actually the replication handler has a bug, which needs to be fixed
>>> - in which case I propose to bump up SOLR-12078 to Blocker.
>>>
>>> I’m open to suggestions.
>>>
>>> —
>>>
>>> Andrzej Białecki
>>>
>>>
>>>
>>
>>
>>
>> --
>> Regards,
>> Shalin Shekhar Mangar.
>>
>>
>
>
>
> --
> Regards,
> Shalin Shekhar Mangar.

-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Shalin Shekhar Mangar]

This is fixed. I committed the fix to master, branch_7x and branch_7_3
branches.

On Wed, Mar 14, 2018 at 9:44 PM, Alan Woodward  wrote:

> Thanks Shalin!
>
>
> On 14 Mar 2018, at 15:50, Shalin Shekhar Mangar 
> wrote:
>
> I'll take a look at it tomorrow morning my time.
>
> On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki <
> andrzej.biale...@lucidworks.com> wrote:
>
>> Well … I looked at it briefly but I have no idea what’s going on there. I
>> could dig into it nonetheless, but if there’s someone who already knows the
>> replication handler ins and outs it would probably get fixed sooner...
>>
>>
>> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
>>
>> I’m happy either way, but if it’s a bug can we get it fixed quickly?  Can
>> you take ownership of this one Andrzej?
>>
>> On 14 Mar 2018, at 11:24, Andrzej Białecki  wrote:
>>
>> Hi,
>>
>> This test has always been fragile, but recently it’s been failing 100%,
>> most often in ‘doTestIndexFetchOnMasterRestart’.
>>
>> I don’t know the replication handler enough to be able to find the real
>> reason behind these failures, but there are two possibilities that I see:
>>
>> * the test has a bug and needs to be fixed - and if we can’t fix it soon
>> then with 7.3 release imminent we could BadApple it until it’s properly
>> fixed
>>
>> * or actually the replication handler has a bug, which needs to be fixed
>> - in which case I propose to bump up SOLR-12078 to Blocker.
>>
>> I’m open to suggestions.
>>
>> —
>>
>> Andrzej Białecki
>>
>>
>>
>>
>
>
> --
> Regards,
> Shalin Shekhar Mangar.
>
>
>


-- 
Regards,
Shalin Shekhar Mangar.

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Alan Woodward]

Thanks Shalin!

> On 14 Mar 2018, at 15:50, Shalin Shekhar Mangar  > wrote:
> 
> I'll take a look at it tomorrow morning my time.
> 
> On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki 
> > 
> wrote:
> Well … I looked at it briefly but I have no idea what’s going on there. I 
> could dig into it nonetheless, but if there’s someone who already knows the 
> replication handler ins and outs it would probably get fixed sooner...
> 
> 
>> On 14 Mar 2018, at 14:23, Alan Woodward > > wrote:
>> 
>> I’m happy either way, but if it’s a bug can we get it fixed quickly?  Can 
>> you take ownership of this one Andrzej?
>> 
>>> On 14 Mar 2018, at 11:24, Andrzej Białecki >> > wrote:
>>> 
>>> Hi,
>>> 
>>> This test has always been fragile, but recently it’s been failing 100%, 
>>> most often in ‘doTestIndexFetchOnMasterRestart’.
>>> 
>>> I don’t know the replication handler enough to be able to find the real 
>>> reason behind these failures, but there are two possibilities that I see:
>>> 
>>> * the test has a bug and needs to be fixed - and if we can’t fix it soon 
>>> then with 7.3 release imminent we could BadApple it until it’s properly 
>>> fixed
>>> 
>>> * or actually the replication handler has a bug, which needs to be fixed - 
>>> in which case I propose to bump up SOLR-12078 to Blocker.
>>> 
>>> I’m open to suggestions.
>>> 
>>> —
>>> 
>>> Andrzej Białecki
>>> 
>> 
> 
> 
> 
> 
> -- 
> Regards,
> Shalin Shekhar Mangar.

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Shalin Shekhar Mangar]

I'll take a look at it tomorrow morning my time.

On Wed, Mar 14, 2018 at 9:07 PM, Andrzej Białecki <
andrzej.biale...@lucidworks.com> wrote:

> Well … I looked at it briefly but I have no idea what’s going on there. I
> could dig into it nonetheless, but if there’s someone who already knows the
> replication handler ins and outs it would probably get fixed sooner...
>
>
> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
>
> I’m happy either way, but if it’s a bug can we get it fixed quickly?  Can
> you take ownership of this one Andrzej?
>
> On 14 Mar 2018, at 11:24, Andrzej Białecki  wrote:
>
> Hi,
>
> This test has always been fragile, but recently it’s been failing 100%,
> most often in ‘doTestIndexFetchOnMasterRestart’.
>
> I don’t know the replication handler enough to be able to find the real
> reason behind these failures, but there are two possibilities that I see:
>
> * the test has a bug and needs to be fixed - and if we can’t fix it soon
> then with 7.3 release imminent we could BadApple it until it’s properly
> fixed
>
> * or actually the replication handler has a bug, which needs to be fixed -
> in which case I propose to bump up SOLR-12078 to Blocker.
>
> I’m open to suggestions.
>
> —
>
> Andrzej Białecki
>
>
>
>


-- 
Regards,
Shalin Shekhar Mangar.

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Andrzej Białecki]

Well … I looked at it briefly but I have no idea what’s going on there. I could 
dig into it nonetheless, but if there’s someone who already knows the 
replication handler ins and outs it would probably get fixed sooner...

> On 14 Mar 2018, at 14:23, Alan Woodward  wrote:
> 
> I’m happy either way, but if it’s a bug can we get it fixed quickly?  Can you 
> take ownership of this one Andrzej?
> 
>> On 14 Mar 2018, at 11:24, Andrzej Białecki > > wrote:
>> 
>> Hi,
>> 
>> This test has always been fragile, but recently it’s been failing 100%, most 
>> often in ‘doTestIndexFetchOnMasterRestart’.
>> 
>> I don’t know the replication handler enough to be able to find the real 
>> reason behind these failures, but there are two possibilities that I see:
>> 
>> * the test has a bug and needs to be fixed - and if we can’t fix it soon 
>> then with 7.3 release imminent we could BadApple it until it’s properly fixed
>> 
>> * or actually the replication handler has a bug, which needs to be fixed - 
>> in which case I propose to bump up SOLR-12078 to Blocker.
>> 
>> I’m open to suggestions.
>> 
>> —
>> 
>> Andrzej Białecki
>> 
> 

Re: TestReplicationHandler is failing 100% (master and 7.x / 7.3)

[Alan Woodward]

I’m happy either way, but if it’s a bug can we get it fixed quickly?  Can you 
take ownership of this one Andrzej?

> On 14 Mar 2018, at 11:24, Andrzej Białecki  > wrote:
> 
> Hi,
> 
> This test has always been fragile, but recently it’s been failing 100%, most 
> often in ‘doTestIndexFetchOnMasterRestart’.
> 
> I don’t know the replication handler enough to be able to find the real 
> reason behind these failures, but there are two possibilities that I see:
> 
> * the test has a bug and needs to be fixed - and if we can’t fix it soon then 
> with 7.3 release imminent we could BadApple it until it’s properly fixed
> 
> * or actually the replication handler has a bug, which needs to be fixed - in 
> which case I propose to bump up SOLR-12078 to Blocker.
> 
> I’m open to suggestions.
> 
> —
> 
> Andrzej Białecki
>