I've posted an hbase 0.20.4 release candidate here:

  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/

Should we release this candidate as hbase 0.20.4?

In includes some critical fixes and some performance enhancements (See
CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
addressed by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Sunday, the 24th of April.

Yours,
The HBase Team

My testing shows it's good (did some extreme PEs and ran unit tests),
although I have some nits WRT documentation.

 - The doc index.html page isn't listing all the new pages that are on
the left, it looks weird.
 - The client package and HTable javadoc are outdated/slightly wrt row
locks since 2248. Here's what Ryan told me after discussing with him:

I dislike the term "respect row lock"  its the wrong model here
Only puts, deletes, check and put, and icv take out row locks now
Gets and scans will return data as fresh and possibly fresher than
when they began and neither will ever return half written data

J-D

On Sun, Apr 18, 2010 at 7:14 AM, Stack <[EMAIL PROTECTED]> wrote:
> I've posted an hbase 0.20.4 release candidate here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>
> Should we release this candidate as hbase 0.20.4?
>
> In includes some critical fixes and some performance enhancements (See
> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
> addressed by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Sunday, the 24th of April.
>
> Yours,
> The HBase Team
>

I've deployed it on 2 clusters and done imports and scans on the tables.
Working good till now.

+1
Amandeep Khurana
Computer Science Graduate Student
University of California, Santa Cruz
On Sat, Apr 17, 2010 at 10:14 PM, Stack <[EMAIL PROTECTED]> wrote:

> I've posted an hbase 0.20.4 release candidate here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>
> Should we release this candidate as hbase 0.20.4?
>
> In includes some critical fixes and some performance enhancements (See
> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
> addressed by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Sunday, the 24th of April.
>
> Yours,
> The HBase Team
>

This rc cannot be official... Todd found a bug and I sent him a
candidate fix. We need to improve the unit tests because they did not
find this but they should have.

On Tuesday, April 20, 2010, Amandeep Khurana <[EMAIL PROTECTED]> wrote:
> I've deployed it on 2 clusters and done imports and scans on the tables.
> Working good till now.
>
> +1
>
>
> Amandeep Khurana
> Computer Science Graduate Student
> University of California, Santa Cruz
>
>
> On Sat, Apr 17, 2010 at 10:14 PM, Stack <[EMAIL PROTECTED]> wrote:
>
>> I've posted an hbase 0.20.4 release candidate here:
>>
>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>>
>> Should we release this candidate as hbase 0.20.4?
>>
>> In includes some critical fixes and some performance enhancements (See
>> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
>> addressed by this candidate).
>>
>> Please take it for a test spin if you have a chance.  Vote closes
>> Sunday, the 24th of April.
>>
>> Yours,
>> The HBase Team
>>
>

On Tue, Apr 20, 2010 at 2:01 AM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
> This rc cannot be official...

Is that a +1 or -1 from you?  What is the RC-sinking issue that you
are referring to?
St.Ack

Todd found a bug and I sent him a
> candidate fix. We need to improve the unit tests because they did not
> find this but they should have.
>
> On Tuesday, April 20, 2010, Amandeep Khurana <[EMAIL PROTECTED]> wrote:
>> I've deployed it on 2 clusters and done imports and scans on the tables.
>> Working good till now.
>>
>> +1
>>
>>
>> Amandeep Khurana
>> Computer Science Graduate Student
>> University of California, Santa Cruz
>>
>>
>> On Sat, Apr 17, 2010 at 10:14 PM, Stack <[EMAIL PROTECTED]> wrote:
>>
>>> I've posted an hbase 0.20.4 release candidate here:
>>>
>>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>>>
>>> Should we release this candidate as hbase 0.20.4?
>>>
>>> In includes some critical fixes and some performance enhancements (See
>>> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
>>> addressed by this candidate).
>>>
>>> Please take it for a test spin if you have a chance.  Vote closes
>>> Sunday, the 24th of April.
>>>
>>> Yours,
>>> The HBase Team
>>>
>>
>

On Tue, Apr 20, 2010 at 1:37 AM, Jean-Daniel Cryans <[EMAIL PROTECTED]> wrote:
> My testing shows it's good (did some extreme PEs and ran unit tests),
> although I have some nits WRT documentation.
>
>  - The doc index.html page isn't listing all the new pages that are on
> the left, it looks weird.
You mean this page:

http://people.apache.org/~stack/hbase-0.20.4-candidate-1/hbase-0.20.4/docs/

What is missing?
>  - The client package and HTable javadoc are outdated/slightly wrt row
> locks since 2248. Here's what Ryan told me after discussing with him:
>
> I dislike the term "respect row lock"  its the wrong model here
> Only puts, deletes, check and put, and icv take out row locks now
> Gets and scans will return data as fresh and possibly fresher than
> when they began and neither will ever return half written data
>
File an issue.

I don't think a doc issue enough to sink the RC but if we cut a new
one, we can update this.

St.Ack
> J-D
>
> On Sun, Apr 18, 2010 at 7:14 AM, Stack <[EMAIL PROTECTED]> wrote:
>> I've posted an hbase 0.20.4 release candidate here:
>>
>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>>
>> Should we release this candidate as hbase 0.20.4?
>>
>> In includes some critical fixes and some performance enhancements (See
>> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
>> addressed by this candidate).
>>
>> Please take it for a test spin if you have a chance.  Vote closes
>> Sunday, the 24th of April.
>>
>> Yours,
>> The HBase Team
>>
>

No JIRA yet but like -1 to the max.

On Apr 20, 2010 8:54 AM, "Stack" <[EMAIL PROTECTED]> wrote:

On Tue, Apr 20, 2010 at 2:01 AM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
> This rc cannot be officia...
Is that a +1 or -1 from you?  What is the RC-sinking issue that you
are referring to?
St.Ack
Todd found a bug and I sent him a
> candidate fix. We need to improve the unit tests because they d...

> You mean this page:
>
> http://people.apache.org/~stack/hbase-0.20.4-candidate-1/hbase-0.20.4/docs/
>
> What is missing?

Well the page has 4 items, on the left I see 9 items. Originally those
two were kept in sync.
>
> File an issue.
>
> I don't think a doc issue enough to sink the RC but if we cut a new
> one, we can update this.

Yep

>
> St.Ack
>
>
>> J-D
>>
>> On Sun, Apr 18, 2010 at 7:14 AM, Stack <[EMAIL PROTECTED]> wrote:
>>> I've posted an hbase 0.20.4 release candidate here:
>>>
>>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>>>
>>> Should we release this candidate as hbase 0.20.4?
>>>
>>> In includes some critical fixes and some performance enhancements (See
>>> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
>>> addressed by this candidate).
>>>
>>> Please take it for a test spin if you have a chance.  Vote closes
>>> Sunday, the 24th of April.
>>>
>>> Yours,
>>> The HBase Team
>>>
>>
>

+1

It made it through a complete benchmark set up on EC2.

   - Andy

Ok, the RC is sunk.

Let's make sure there is a unit test for whatever the issue found is. The suite otherwise didn't catch it and neither did PE, at least not an obvious failure.

Hey Stack, anything else that is worth pulling into the next RC in addition to the doc updates suggested by J-D? For example HBASE-2421, HBASE-2424, HBASE-2441, HBASE-2443?

   - Andy

--- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
> No JIRA yet but like -1 to the max.
[...]
> Todd found a bug and I sent him a candidate fix. We need to improve
> the unit tests because they d...

On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell <[EMAIL PROTECTED]> wrote:

> Ok, the RC is sunk.
>
> Let's make sure there is a unit test for whatever the issue found is. The
> suite otherwise didn't catch it and neither did PE, at least not an obvious
> failure.
>

Yep, I sent Ryan a functional test which he's now posted to the JIRA - it
doesn't currently run against a minicluster but it should be trivial. (built
the test as part of the suite I'm doing for testing all of our acid
guarantees)
>
> Hey Stack, anything else that is worth pulling into the next RC in addition
> to the doc updates suggested by J-D? For example HBASE-2421, HBASE-2424,
> HBASE-2441, HBASE-2443?
>
>   - Andy
>
> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
> > No JIRA yet but like -1 to the max.
> [...]
> > Todd found a bug and I sent him a candidate fix. We need to improve
> > the unit tests because they d...
>
>
>
>
>
>
--
Todd Lipcon
Software Engineer, Cloudera

Here is the JIRA:

https://issues.apache.org/jira/browse/HBASE-2474

I think a trivial fix to the existing unit tests can cover this case,
since they are validating the return count, but not the contents.
Just add in a bit of content validation and we should be good.

-ryan

On Tue, Apr 20, 2010 at 2:48 PM, Todd Lipcon <[EMAIL PROTECTED]> wrote:
> On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell <[EMAIL PROTECTED]> wrote:
>
>> Ok, the RC is sunk.
>>
>> Let's make sure there is a unit test for whatever the issue found is. The
>> suite otherwise didn't catch it and neither did PE, at least not an obvious
>> failure.
>>
>
> Yep, I sent Ryan a functional test which he's now posted to the JIRA - it
> doesn't currently run against a minicluster but it should be trivial. (built
> the test as part of the suite I'm doing for testing all of our acid
> guarantees)
>
>
>>
>> Hey Stack, anything else that is worth pulling into the next RC in addition
>> to the doc updates suggested by J-D? For example HBASE-2421, HBASE-2424,
>> HBASE-2441, HBASE-2443?
>>
>>   - Andy
>>
>> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>> > No JIRA yet but like -1 to the max.
>> [...]
>> > Todd found a bug and I sent him a candidate fix. We need to improve
>> > the unit tests because they d...
>>
>>
>>
>>
>>
>>
>
>
> --
> Todd Lipcon
> Software Engineer, Cloudera
>

That explains why PE was useless also in this case.

> Here is the JIRA:
>
> https://issues.apache.org/jira/browse/HBASE-2474

HBASE-2421, 2424 are multiput fixes.  Multiput is not in
0.20_pre_durability.   The others I'd like to do unit tests for first
and IMO, they are ugly but not blockers.

I'd suggest that as soon as Ryan has a fix for 2474, we roll a new RC.
 J-D, you have a euro-fix for the doc issue?

St.Ack

On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell <[EMAIL PROTECTED]> wrote:
> Ok, the RC is sunk.
>
> Let's make sure there is a unit test for whatever the issue found is. The suite otherwise didn't catch it and neither did PE, at least not an obvious failure.
>
> Hey Stack, anything else that is worth pulling into the next RC in addition to the doc updates suggested by J-D? For example HBASE-2421, HBASE-2424, HBASE-2441, HBASE-2443?
>
>   - Andy
>
> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>> No JIRA yet but like -1 to the max.
> [...]
>> Todd found a bug and I sent him a candidate fix. We need to improve
>> the unit tests because they d...
>
>
>
>
>
>

Ok 2474 is committed to both 0.20 and 0.20_pre_durability

I still havent quite ported 2248 to trunk yet, it isnt that simple
because it depends on a few other patches which I need to check.  Once
we have 0.20.4 released I'll work on it more.

On Tue, Apr 20, 2010 at 3:07 PM, Stack <[EMAIL PROTECTED]> wrote:
> HBASE-2421, 2424 are multiput fixes.  Multiput is not in
> 0.20_pre_durability.   The others I'd like to do unit tests for first
> and IMO, they are ugly but not blockers.
>
> I'd suggest that as soon as Ryan has a fix for 2474, we roll a new RC.
>  J-D, you have a euro-fix for the doc issue?
>
> St.Ack
>
> On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell <[EMAIL PROTECTED]> wrote:
>> Ok, the RC is sunk.
>>
>> Let's make sure there is a unit test for whatever the issue found is. The suite otherwise didn't catch it and neither did PE, at least not an obvious failure.
>>
>> Hey Stack, anything else that is worth pulling into the next RC in addition to the doc updates suggested by J-D? For example HBASE-2421, HBASE-2424, HBASE-2441, HBASE-2443?
>>
>>   - Andy
>>
>> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>>> No JIRA yet but like -1 to the max.
>> [...]
>>> Todd found a bug and I sent him a candidate fix. We need to improve
>>> the unit tests because they d...
>>
>>
>>
>>
>>
>>
>

Do you want to make a quick javadoc fix for scan and locks?

On Apr 20, 2010, at 5:21 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:

> Ok 2474 is committed to both 0.20 and 0.20_pre_durability
>
> I still havent quite ported 2248 to trunk yet, it isnt that simple
> because it depends on a few other patches which I need to check.  Once
> we have 0.20.4 released I'll work on it more.
>
> On Tue, Apr 20, 2010 at 3:07 PM, Stack <[EMAIL PROTECTED]> wrote:
>> HBASE-2421, 2424 are multiput fixes.  Multiput is not in
>> 0.20_pre_durability.   The others I'd like to do unit tests for first
>> and IMO, they are ugly but not blockers.
>>
>> I'd suggest that as soon as Ryan has a fix for 2474, we roll a new  
>> RC.
>>  J-D, you have a euro-fix for the doc issue?
>>
>> St.Ack
>>
>> On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell  
>> <[EMAIL PROTECTED]> wrote:
>>> Ok, the RC is sunk.
>>>
>>> Let's make sure there is a unit test for whatever the issue found  
>>> is. The suite otherwise didn't catch it and neither did PE, at  
>>> least not an obvious failure.
>>>
>>> Hey Stack, anything else that is worth pulling into the next RC in  
>>> addition to the doc updates suggested by J-D? For example  
>>> HBASE-2421, HBASE-2424, HBASE-2441, HBASE-2443?
>>>
>>>   - Andy
>>>
>>> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>>>> No JIRA yet but like -1 to the max.
>>> [...]
>>>> Todd found a bug and I sent him a candidate fix. We need to improve
>>>> the unit tests because they d...
>>>
>>>
>>>
>>>
>>>
>>>
>>

Nick said he had rewritten the javadoc for a post 2248 world... but i
havent seen him online.

i can do a quick fix

On Tue, Apr 20, 2010 at 5:26 PM, Stack <[EMAIL PROTECTED]> wrote:
> Do you want to make a quick javadoc fix for scan and locks?
>
>
>
> On Apr 20, 2010, at 5:21 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>
>> Ok 2474 is committed to both 0.20 and 0.20_pre_durability
>>
>> I still havent quite ported 2248 to trunk yet, it isnt that simple
>> because it depends on a few other patches which I need to check.  Once
>> we have 0.20.4 released I'll work on it more.
>>
>> On Tue, Apr 20, 2010 at 3:07 PM, Stack <[EMAIL PROTECTED]> wrote:
>>>
>>> HBASE-2421, 2424 are multiput fixes.  Multiput is not in
>>> 0.20_pre_durability.   The others I'd like to do unit tests for first
>>> and IMO, they are ugly but not blockers.
>>>
>>> I'd suggest that as soon as Ryan has a fix for 2474, we roll a new RC.
>>>  J-D, you have a euro-fix for the doc issue?
>>>
>>> St.Ack
>>>
>>> On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell <[EMAIL PROTECTED]>
>>> wrote:
>>>>
>>>> Ok, the RC is sunk.
>>>>
>>>> Let's make sure there is a unit test for whatever the issue found is.
>>>> The suite otherwise didn't catch it and neither did PE, at least not an
>>>> obvious failure.
>>>>
>>>> Hey Stack, anything else that is worth pulling into the next RC in
>>>> addition to the doc updates suggested by J-D? For example HBASE-2421,
>>>> HBASE-2424, HBASE-2441, HBASE-2443?
>>>>
>>>>  - Andy
>>>>
>>>> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>>>>>
>>>>> No JIRA yet but like -1 to the max.
>>>>
>>>> [...]
>>>>>
>>>>> Todd found a bug and I sent him a candidate fix. We need to improve
>>>>> the unit tests because they d...
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>

Ok I fixed the javadoc in both branches.

-ryan

On Tue, Apr 20, 2010 at 5:28 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
> Nick said he had rewritten the javadoc for a post 2248 world... but i
> havent seen him online.
>
> i can do a quick fix
>
> On Tue, Apr 20, 2010 at 5:26 PM, Stack <[EMAIL PROTECTED]> wrote:
>> Do you want to make a quick javadoc fix for scan and locks?
>>
>>
>>
>> On Apr 20, 2010, at 5:21 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>>
>>> Ok 2474 is committed to both 0.20 and 0.20_pre_durability
>>>
>>> I still havent quite ported 2248 to trunk yet, it isnt that simple
>>> because it depends on a few other patches which I need to check.  Once
>>> we have 0.20.4 released I'll work on it more.
>>>
>>> On Tue, Apr 20, 2010 at 3:07 PM, Stack <[EMAIL PROTECTED]> wrote:
>>>>
>>>> HBASE-2421, 2424 are multiput fixes.  Multiput is not in
>>>> 0.20_pre_durability.   The others I'd like to do unit tests for first
>>>> and IMO, they are ugly but not blockers.
>>>>
>>>> I'd suggest that as soon as Ryan has a fix for 2474, we roll a new RC.
>>>>  J-D, you have a euro-fix for the doc issue?
>>>>
>>>> St.Ack
>>>>
>>>> On Tue, Apr 20, 2010 at 2:46 PM, Andrew Purtell <[EMAIL PROTECTED]>
>>>> wrote:
>>>>>
>>>>> Ok, the RC is sunk.
>>>>>
>>>>> Let's make sure there is a unit test for whatever the issue found is.
>>>>> The suite otherwise didn't catch it and neither did PE, at least not an
>>>>> obvious failure.
>>>>>
>>>>> Hey Stack, anything else that is worth pulling into the next RC in
>>>>> addition to the doc updates suggested by J-D? For example HBASE-2421,
>>>>> HBASE-2424, HBASE-2441, HBASE-2443?
>>>>>
>>>>>  - Andy
>>>>>
>>>>> --- On Tue, 4/20/10, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>>>>>>
>>>>>> No JIRA yet but like -1 to the max.
>>>>>
>>>>> [...]
>>>>>>
>>>>>> Todd found a bug and I sent him a candidate fix. We need to improve
>>>>>> the unit tests because they d...
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>
>

On Tue, Apr 20, 2010 at 9:02 AM, Jean-Daniel Cryans <[EMAIL PROTECTED]> wrote:
>> You mean this page:
>>
>> http://people.apache.org/~stack/hbase-0.20.4-candidate-1/hbase-0.20.4/docs/
>>
>> What is missing?
>
> Well the page has 4 items, on the left I see 9 items. Originally those
> two were kept in sync.
>
Oh, you are saying that the navbar and the index page should have same
number of items?  That what is in the navbar should be here in the
main index page for the doc?

I don't think those were ever in sync; e.g. we didn't have link to
overview and mailing lists in main page.

Hope you don't mind that I go ahead an roll an RC2, given that the
javadoc and  Ryans' 2248 bug have been fixed, while you are sleeping
in your Lucerne bed.......

St.Ack
>
>>
>> File an issue.
>>
>> I don't think a doc issue enough to sink the RC but if we cut a new
>> one, we can update this.
>
> Yep
>
>>
>> St.Ack
>>
>>
>>> J-D
>>>
>>> On Sun, Apr 18, 2010 at 7:14 AM, Stack <[EMAIL PROTECTED]> wrote:
>>>> I've posted an hbase 0.20.4 release candidate here:
>>>>
>>>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>>>>
>>>> Should we release this candidate as hbase 0.20.4?
>>>>
>>>> In includes some critical fixes and some performance enhancements (See
>>>> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
>>>> addressed by this candidate).
>>>>
>>>> Please take it for a test spin if you have a chance.  Vote closes
>>>> Sunday, the 24th of April.
>>>>
>>>> Yours,
>>>> The HBase Team
>>>>
>>>
>>
>

I've posted an hbase 0.20.4 release candidate 2 here:

  http://people.apache.org/~stack/hbase-0.20.4-candidate-2/

Should we release this candidate as hbase 0.20.4?

In includes some critical fixes and some performance enhancements (See
CHANGES.txt or the release notes in the docs [1] for the list of over
100 issues addressed by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Tuesday, the 27th of April.

Yours,
The HBase Team

1. http://people.apache.org/~stack/hbase-0.20.4-candidate-2/hbase-0.20.4/docs/releasenotes.html
On Sat, Apr 17, 2010 at 10:14 PM, Stack <[EMAIL PROTECTED]> wrote:
> I've posted an hbase 0.20.4 release candidate here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-1/
>
> Should we release this candidate as hbase 0.20.4?
>
> In includes some critical fixes and some performance enhancements (See
> CHANGES.txt or http://su.pr/4QEGFv for the list of over 100 issues
> addressed by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Sunday, the 24th of April.
>
> Yours,
> The HBase Team
>

I've posted an hbase 0.20.4 release candidate 3 here:

  http://people.apache.org/~stack/hbase-0.20.4-candidate-3/

Should we release this candidate as hbase 0.20.4?

It includes some critical fixes and some performance enhancements (See
CHANGES.txt or the release notes in the docs [1] for the list of 112
issues addressed by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Thursday, the 29th of April.

Yours,
Your HBase Team

1. https://issues.apache.org/jira/browse/HBASE/fixforversion/12314496

Not sure if it's a blocker, but HBASE-2378 doesn't seem to be in the
pre_durability branch, though the HTML version of the release notes says it
is (it's not in CHANGES.txt either)

(I know we have at least a couple customers for whom HFileOutputFormat is
useless without this patch)

I'll give the rc a go on my test cluster this week.

-Todd
On Sat, Apr 24, 2010 at 12:40 PM, Stack <[EMAIL PROTECTED]> wrote:

> I've posted an hbase 0.20.4 release candidate 3 here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-3/
>
> Should we release this candidate as hbase 0.20.4?
>
> It includes some critical fixes and some performance enhancements (See
> CHANGES.txt or the release notes in the docs [1] for the list of 112
> issues addressed by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Thursday, the 29th of April.
>
> Yours,
> Your HBase Team
>
> 1. https://issues.apache.org/jira/browse/HBASE/fixforversion/12314496
>

--
Todd Lipcon
Software Engineer, Cloudera

I've posted an hbase 0.20.4 release candidate 4 here:

  http://people.apache.org/~stack/hbase-0.20.4-candidate-3/

Should we release this candidate as hbase 0.20.4?

It includes some critical fixes and some performance enhancements (See
CHANGES.txt or the release notes in the docs [1] for issues addressed
by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Friday, the 30th of April.

Yours,
Your HBase Team

1. http://people.apache.org/~stack/hbase-0.20.4-candidate-4/hbase-0.20.4/docs/releasenotes.html

correct link seems to be
http://people.apache.org/~stack/hbase-0.20.4-candidate-4/

s/3/4/ :-)
Stack:
> I've posted an hbase 0.20.4 release candidate 4 here:
>
>   http://people.apache.org/~stack/hbase-0.20.4-candidate-3/
>
> Should we release this candidate as hbase 0.20.4?
>
> It includes some critical fixes and some performance enhancements (See
> CHANGES.txt or the release notes in the docs [1] for issues addressed
> by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Friday, the 30th of April.
>
> Yours,
> Your HBase Team
>
> 1.
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-4/hbase-0.20.4/docs
> /releasenotes.html
>

Thomas Koch, http://www.koch.ro

-1 https://issues.apache.org/jira/browse/HBASE-2499

I thought we fix it, but we didn't. It pretty much prevents anyone
from altering a largish table.

J-D

On Mon, Apr 26, 2010 at 10:55 PM, Stack <[EMAIL PROTECTED]> wrote:
> I've posted an hbase 0.20.4 release candidate 4 here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-3/
>
> Should we release this candidate as hbase 0.20.4?
>
> It includes some critical fixes and some performance enhancements (See
> CHANGES.txt or the release notes in the docs [1] for issues addressed
> by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Friday, the 30th of April.
>
> Yours,
> Your HBase Team
>
> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-4/hbase-0.20.4/docs/releasenotes.html
>

I agree.  New RC coming now.
Thanks J-D,
St.Ack

On Wed, Apr 28, 2010 at 3:12 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]> wrote:
> -1 https://issues.apache.org/jira/browse/HBASE-2499
>
> I thought we fix it, but we didn't. It pretty much prevents anyone
> from altering a largish table.
>
> J-D
>
> On Mon, Apr 26, 2010 at 10:55 PM, Stack <[EMAIL PROTECTED]> wrote:
>> I've posted an hbase 0.20.4 release candidate 4 here:
>>
>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-3/
>>
>> Should we release this candidate as hbase 0.20.4?
>>
>> It includes some critical fixes and some performance enhancements (See
>> CHANGES.txt or the release notes in the docs [1] for issues addressed
>> by this candidate).
>>
>> Please take it for a test spin if you have a chance.  Vote closes
>> Friday, the 30th of April.
>>
>> Yours,
>> Your HBase Team
>>
>> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-4/hbase-0.20.4/docs/releasenotes.html
>>
>

I've posted an hbase 0.20.4 release candidate 5 here:

 http://people.apache.org/~stack/hbase-0.20.4-candidate-5/

It includes a fix for the issue J-D found offlining big tables.

Should we release this candidate as hbase 0.20.4?

It includes critical fixes and some performance enhancements (See
CHANGES.txt or the release notes in the docs [1] for issues addressed
by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Saturday, the 1st of May.

Yours,
Your HBase Team

1. http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html

+1: doc is good, passes unit tests, passes my own stress testing.

J-D

On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
> I've posted an hbase 0.20.4 release candidate 5 here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
>
> It includes a fix for the issue J-D found offlining big tables.
>
> Should we release this candidate as hbase 0.20.4?
>
> It includes critical fixes and some performance enhancements (See
> CHANGES.txt or the release notes in the docs [1] for issues addressed
> by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Saturday, the 1st of May.
>
> Yours,
> Your HBase Team
>
> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
>

+1, I ran a small data load overnight of about 1.2B rows (~88GB) on my 5
node cluster. Grepped logs for anything suspicious looking, everything looks
good.

-Todd

On Thu, Apr 29, 2010 at 9:00 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:

> +1: doc is good, passes unit tests, passes my own stress testing.
>
> J-D
>
> On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
> > I've posted an hbase 0.20.4 release candidate 5 here:
> >
> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
> >
> > It includes a fix for the issue J-D found offlining big tables.
> >
> > Should we release this candidate as hbase 0.20.4?
> >
> > It includes critical fixes and some performance enhancements (See
> > CHANGES.txt or the release notes in the docs [1] for issues addressed
> > by this candidate).
> >
> > Please take it for a test spin if you have a chance.  Vote closes
> > Saturday, the 1st of May.
> >
> > Yours,
> > Your HBase Team
> >
> > 1.
> http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
> >
>

--
Todd Lipcon
Software Engineer, Cloudera

+1

Loaded 1B using PE into 5node cluster.  Did scans and reads.  No
noticeable issues.  Downloaded it. Made sure it runs in local mode.
Ran PE against it (I think the ZK disconnect WARN message cleaned up
in zk 3.3).  Doc looks fine.

St.Ack

On Fri, Apr 30, 2010 at 11:57 AM, Todd Lipcon <[EMAIL PROTECTED]> wrote:
> +1, I ran a small data load overnight of about 1.2B rows (~88GB) on my 5
> node cluster. Grepped logs for anything suspicious looking, everything looks
> good.
>
> -Todd
>
> On Thu, Apr 29, 2010 at 9:00 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:
>
>> +1: doc is good, passes unit tests, passes my own stress testing.
>>
>> J-D
>>
>> On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
>> > I've posted an hbase 0.20.4 release candidate 5 here:
>> >
>> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
>> >
>> > It includes a fix for the issue J-D found offlining big tables.
>> >
>> > Should we release this candidate as hbase 0.20.4?
>> >
>> > It includes critical fixes and some performance enhancements (See
>> > CHANGES.txt or the release notes in the docs [1] for issues addressed
>> > by this candidate).
>> >
>> > Please take it for a test spin if you have a chance.  Vote closes
>> > Saturday, the 1st of May.
>> >
>> > Yours,
>> > Your HBase Team
>> >
>> > 1.
>> http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
>> >
>>
>
>
>
> --
> Todd Lipcon
> Software Engineer, Cloudera
>

I used the Import mapreduce to pull in 16GB of data and things looked fine.

-ryan

On Sun, May 2, 2010 at 9:34 AM, Stack <[EMAIL PROTECTED]> wrote:
> +1
>
> Loaded 1B using PE into 5node cluster.  Did scans and reads.  No
> noticeable issues.  Downloaded it. Made sure it runs in local mode.
> Ran PE against it (I think the ZK disconnect WARN message cleaned up
> in zk 3.3).  Doc looks fine.
>
> St.Ack
>
> On Fri, Apr 30, 2010 at 11:57 AM, Todd Lipcon <[EMAIL PROTECTED]> wrote:
>> +1, I ran a small data load overnight of about 1.2B rows (~88GB) on my 5
>> node cluster. Grepped logs for anything suspicious looking, everything looks
>> good.
>>
>> -Todd
>>
>> On Thu, Apr 29, 2010 at 9:00 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:
>>
>>> +1: doc is good, passes unit tests, passes my own stress testing.
>>>
>>> J-D
>>>
>>> On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
>>> > I've posted an hbase 0.20.4 release candidate 5 here:
>>> >
>>> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
>>> >
>>> > It includes a fix for the issue J-D found offlining big tables.
>>> >
>>> > Should we release this candidate as hbase 0.20.4?
>>> >
>>> > It includes critical fixes and some performance enhancements (See
>>> > CHANGES.txt or the release notes in the docs [1] for issues addressed
>>> > by this candidate).
>>> >
>>> > Please take it for a test spin if you have a chance.  Vote closes
>>> > Saturday, the 1st of May.
>>> >
>>> > Yours,
>>> > Your HBase Team
>>> >
>>> > 1.
>>> http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
>>> >
>>>
>>
>>
>>
>> --
>> Todd Lipcon
>> Software Engineer, Cloudera
>>
>

VOTE PASSES because 4x+1 && !-1s
Let me publish RC5 as 0.20.4.
Thanks to all who tested,
St.Ack

On Mon, May 3, 2010 at 1:24 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
> I used the Import mapreduce to pull in 16GB of data and things looked fine.
>
> -ryan
>
> On Sun, May 2, 2010 at 9:34 AM, Stack <[EMAIL PROTECTED]> wrote:
>> +1
>>
>> Loaded 1B using PE into 5node cluster.  Did scans and reads.  No
>> noticeable issues.  Downloaded it. Made sure it runs in local mode.
>> Ran PE against it (I think the ZK disconnect WARN message cleaned up
>> in zk 3.3).  Doc looks fine.
>>
>> St.Ack
>>
>> On Fri, Apr 30, 2010 at 11:57 AM, Todd Lipcon <[EMAIL PROTECTED]> wrote:
>>> +1, I ran a small data load overnight of about 1.2B rows (~88GB) on my 5
>>> node cluster. Grepped logs for anything suspicious looking, everything looks
>>> good.
>>>
>>> -Todd
>>>
>>> On Thu, Apr 29, 2010 at 9:00 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:
>>>
>>>> +1: doc is good, passes unit tests, passes my own stress testing.
>>>>
>>>> J-D
>>>>
>>>> On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
>>>> > I've posted an hbase 0.20.4 release candidate 5 here:
>>>> >
>>>> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
>>>> >
>>>> > It includes a fix for the issue J-D found offlining big tables.
>>>> >
>>>> > Should we release this candidate as hbase 0.20.4?
>>>> >
>>>> > It includes critical fixes and some performance enhancements (See
>>>> > CHANGES.txt or the release notes in the docs [1] for issues addressed
>>>> > by this candidate).
>>>> >
>>>> > Please take it for a test spin if you have a chance.  Vote closes
>>>> > Saturday, the 1st of May.
>>>> >
>>>> > Yours,
>>>> > Your HBase Team
>>>> >
>>>> > 1.
>>>> http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
>>>> >
>>>>
>>>
>>>
>>>
>>> --
>>> Todd Lipcon
>>> Software Engineer, Cloudera
>>>
>>
>

Hurray!

On Mon, May 3, 2010 at 3:06 PM, Stack <[EMAIL PROTECTED]> wrote:
> VOTE PASSES because 4x+1 && !-1s
> Let me publish RC5 as 0.20.4.
> Thanks to all who tested,
> St.Ack
>
> On Mon, May 3, 2010 at 1:24 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>> I used the Import mapreduce to pull in 16GB of data and things looked fine.
>>
>> -ryan
>>
>> On Sun, May 2, 2010 at 9:34 AM, Stack <[EMAIL PROTECTED]> wrote:
>>> +1
>>>
>>> Loaded 1B using PE into 5node cluster.  Did scans and reads.  No
>>> noticeable issues.  Downloaded it. Made sure it runs in local mode.
>>> Ran PE against it (I think the ZK disconnect WARN message cleaned up
>>> in zk 3.3).  Doc looks fine.
>>>
>>> St.Ack
>>>
>>> On Fri, Apr 30, 2010 at 11:57 AM, Todd Lipcon <[EMAIL PROTECTED]> wrote:
>>>> +1, I ran a small data load overnight of about 1.2B rows (~88GB) on my 5
>>>> node cluster. Grepped logs for anything suspicious looking, everything looks
>>>> good.
>>>>
>>>> -Todd
>>>>
>>>> On Thu, Apr 29, 2010 at 9:00 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:
>>>>
>>>>> +1: doc is good, passes unit tests, passes my own stress testing.
>>>>>
>>>>> J-D
>>>>>
>>>>> On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
>>>>> > I've posted an hbase 0.20.4 release candidate 5 here:
>>>>> >
>>>>> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
>>>>> >
>>>>> > It includes a fix for the issue J-D found offlining big tables.
>>>>> >
>>>>> > Should we release this candidate as hbase 0.20.4?
>>>>> >
>>>>> > It includes critical fixes and some performance enhancements (See
>>>>> > CHANGES.txt or the release notes in the docs [1] for issues addressed
>>>>> > by this candidate).
>>>>> >
>>>>> > Please take it for a test spin if you have a chance.  Vote closes
>>>>> > Saturday, the 1st of May.
>>>>> >
>>>>> > Yours,
>>>>> > Your HBase Team
>>>>> >
>>>>> > 1.
>>>>> http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
>>>>> >
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Todd Lipcon
>>>> Software Engineer, Cloudera
>>>>
>>>
>>
>

Cancel that!

We just tripped over HBASE-2509.

Will cut a new RC after we have a fix.

St.Ack

On Mon, May 3, 2010 at 3:06 PM, Stack <[EMAIL PROTECTED]> wrote:
> VOTE PASSES because 4x+1 && !-1s
> Let me publish RC5 as 0.20.4.
> Thanks to all who tested,
> St.Ack
>
> On Mon, May 3, 2010 at 1:24 PM, Ryan Rawson <[EMAIL PROTECTED]> wrote:
>> I used the Import mapreduce to pull in 16GB of data and things looked fine.
>>
>> -ryan
>>
>> On Sun, May 2, 2010 at 9:34 AM, Stack <[EMAIL PROTECTED]> wrote:
>>> +1
>>>
>>> Loaded 1B using PE into 5node cluster.  Did scans and reads.  No
>>> noticeable issues.  Downloaded it. Made sure it runs in local mode.
>>> Ran PE against it (I think the ZK disconnect WARN message cleaned up
>>> in zk 3.3).  Doc looks fine.
>>>
>>> St.Ack
>>>
>>> On Fri, Apr 30, 2010 at 11:57 AM, Todd Lipcon <[EMAIL PROTECTED]> wrote:
>>>> +1, I ran a small data load overnight of about 1.2B rows (~88GB) on my 5
>>>> node cluster. Grepped logs for anything suspicious looking, everything looks
>>>> good.
>>>>
>>>> -Todd
>>>>
>>>> On Thu, Apr 29, 2010 at 9:00 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]>wrote:
>>>>
>>>>> +1: doc is good, passes unit tests, passes my own stress testing.
>>>>>
>>>>> J-D
>>>>>
>>>>> On Wed, Apr 28, 2010 at 10:59 PM, Stack <[EMAIL PROTECTED]> wrote:
>>>>> > I've posted an hbase 0.20.4 release candidate 5 here:
>>>>> >
>>>>> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-5/
>>>>> >
>>>>> > It includes a fix for the issue J-D found offlining big tables.
>>>>> >
>>>>> > Should we release this candidate as hbase 0.20.4?
>>>>> >
>>>>> > It includes critical fixes and some performance enhancements (See
>>>>> > CHANGES.txt or the release notes in the docs [1] for issues addressed
>>>>> > by this candidate).
>>>>> >
>>>>> > Please take it for a test spin if you have a chance.  Vote closes
>>>>> > Saturday, the 1st of May.
>>>>> >
>>>>> > Yours,
>>>>> > Your HBase Team
>>>>> >
>>>>> > 1.
>>>>> http://people.apache.org/~stack/hbase-0.20.4-candidate-5/hbase-0.20.4/docs/releasenotes.html
>>>>> >
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Todd Lipcon
>>>> Software Engineer, Cloudera
>>>>
>>>
>>
>

I've posted an hbase 0.20.4 release candidate 6 here:

 http://people.apache.org/~stack/hbase-0.20.4-candidate-6/

Should we release this candidate as hbase 0.20.4?

It includes some critical fixes and some performance enhancements (See
CHANGES.txt or the release notes in the docs [1] for issues addressed
by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Thursday, the 6th of May.  RC6 has one fix over RC5 (hbase-2509),
hence the short vote period.

Yours,
Your HBase Team

1. http://people.apache.org/~stack/hbase-0.20.4-candidate-6/hbase-0.20.4/docs/releasenotes.html

I am +1, as the person who authored the fix (but not the test, thanks stack!).

It's a very small fix for a rare, but annoying condition.

-ryan

On Tue, May 4, 2010 at 4:52 PM, Stack <[EMAIL PROTECTED]> wrote:
> I've posted an hbase 0.20.4 release candidate 6 here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
>
> Should we release this candidate as hbase 0.20.4?
>
> It includes some critical fixes and some performance enhancements (See
> CHANGES.txt or the release notes in the docs [1] for issues addressed
> by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Thursday, the 6th of May.  RC6 has one fix over RC5 (hbase-2509),
> hence the short vote period.
>
> Yours,
> Your HBase Team
>
> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-6/hbase-0.20.4/docs/releasenotes.html
>

Did some more random read-oriented testing and everything is fine. +1

J-D

On Tue, May 4, 2010 at 4:52 PM, Stack <[EMAIL PROTECTED]> wrote:
> I've posted an hbase 0.20.4 release candidate 6 here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
>
> Should we release this candidate as hbase 0.20.4?
>
> It includes some critical fixes and some performance enhancements (See
> CHANGES.txt or the release notes in the docs [1] for issues addressed
> by this candidate).
>
> Please take it for a test spin if you have a chance.  Vote closes
> Thursday, the 6th of May.  RC6 has one fix over RC5 (hbase-2509),
> hence the short vote period.
>
> Yours,
> Your HBase Team
>
> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-6/hbase-0.20.4/docs/releasenotes.html
>

All test pass for me, no real load test though.

+1

On Wed, May 5, 2010 at 4:50 AM, Jean-Daniel Cryans <[EMAIL PROTECTED]> wrote:
> Did some more random read-oriented testing and everything is fine. +1
>
> J-D
>
> On Tue, May 4, 2010 at 4:52 PM, Stack <[EMAIL PROTECTED]> wrote:
>> I've posted an hbase 0.20.4 release candidate 6 here:
>>
>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
>>
>> Should we release this candidate as hbase 0.20.4?
>>
>> It includes some critical fixes and some performance enhancements (See
>> CHANGES.txt or the release notes in the docs [1] for issues addressed
>> by this candidate).
>>
>> Please take it for a test spin if you have a chance.  Vote closes
>> Thursday, the 6th of May.  RC6 has one fix over RC5 (hbase-2509),
>> hence the short vote period.
>>
>> Yours,
>> Your HBase Team
>>
>> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-6/hbase-0.20.4/docs/releasenotes.html
>>
>

I'm +1 too.  Ran YCSB w/ mixed loading over night.  Found hbase-2516
but thats an old issue made a little more acute now Gets are Scans.

Let me push out the release.  4+1s.

St.Ack
On Tue, May 4, 2010 at 7:50 PM, Jean-Daniel Cryans <[EMAIL PROTECTED]> wrote:
> Did some more random read-oriented testing and everything is fine. +1
>
> J-D
>
> On Tue, May 4, 2010 at 4:52 PM, Stack <[EMAIL PROTECTED]> wrote:
>> I've posted an hbase 0.20.4 release candidate 6 here:
>>
>>  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
>>
>> Should we release this candidate as hbase 0.20.4?
>>
>> It includes some critical fixes and some performance enhancements (See
>> CHANGES.txt or the release notes in the docs [1] for issues addressed
>> by this candidate).
>>
>> Please take it for a test spin if you have a chance.  Vote closes
>> Thursday, the 6th of May.  RC6 has one fix over RC5 (hbase-2509),
>> hence the short vote period.
>>
>> Yours,
>> Your HBase Team
>>
>> 1. http://people.apache.org/~stack/hbase-0.20.4-candidate-6/hbase-0.20.4/docs/releasenotes.html
>>
>

Do I understand correctly that HBASE-2515 means we will lose regions (stuck in transition) if a table is disabled at an inopportune time after a split? If so, -1, let's pull HBASE-2515 into 0.20.4. This may have been there for a while but now it's a known problem. It's still necessary to disable tables to make schema changes.

   - Andy

> From: Stack <[EMAIL PROTECTED]>
> Subject: [VOTE] Shall we release hbase-0.20.4 Release Candidate 6 as  hbase-0.20.4?
> To: "HBase Dev List" <[EMAIL PROTECTED]>
> Date: Tuesday, May 4, 2010, 4:52 PM
> I've posted an hbase 0.20.4 release
> candidate 6 here:
>
>  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
>
> Should we release this candidate as hbase 0.20.4?

I missed that the vote closed early. Disregard the -1 as it came too late.

  - Andy

> From: Andrew Purtell <[EMAIL PROTECTED]>
> Subject: Re: [VOTE] Shall we release hbase-0.20.4 Release Candidate 6 as  hbase-0.20.4?
> To: [EMAIL PROTECTED]
> Date: Wednesday, May 5, 2010, 12:55 PM
> Do I understand correctly that
> HBASE-2515 means we will lose regions (stuck in transition)
> if a table is disabled at an inopportune time after a split?
> If so, -1, let's pull HBASE-2515 into 0.20.4. This may have
> been there for a while but now it's a known problem. It's
> still necessary to disable tables to make schema changes.
>
>    - Andy
>
> > From: Stack <[EMAIL PROTECTED]>
> > Subject: [VOTE] Shall we release hbase-0.20.4 Release
> Candidate 6 as  hbase-0.20.4?
> > To: "HBase Dev List" <[EMAIL PROTECTED]>
> > Date: Tuesday, May 4, 2010, 4:52 PM
> > I've posted an hbase 0.20.4 release
> > candidate 6 here:
> >
> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
> >
> > Should we release this candidate as hbase 0.20.4?
>
>
>
>      
>
>

About the issue you were raising, it's a bit different than what you described.

If you try disabling a table that still has split parents in .META.,
those parents will be put in the regionsInTransitions list and marked
as closing... but they won't be closed since they are offline already.
It's when you re-enable that table that it may mess things up a bit,
nothing a restart can't fix (I didn't dig that far in my testing).

J-D

On Wed, May 5, 2010 at 1:10 PM, Andrew Purtell <[EMAIL PROTECTED]> wrote:
> I missed that the vote closed early. Disregard the -1 as it came too late.
>
>  - Andy
>
>> From: Andrew Purtell <[EMAIL PROTECTED]>
>> Subject: Re: [VOTE] Shall we release hbase-0.20.4 Release Candidate 6 as  hbase-0.20.4?
>> To: [EMAIL PROTECTED]
>> Date: Wednesday, May 5, 2010, 12:55 PM
>> Do I understand correctly that
>> HBASE-2515 means we will lose regions (stuck in transition)
>> if a table is disabled at an inopportune time after a split?
>> If so, -1, let's pull HBASE-2515 into 0.20.4. This may have
>> been there for a while but now it's a known problem. It's
>> still necessary to disable tables to make schema changes.
>>
>>    - Andy
>>
>> > From: Stack <[EMAIL PROTECTED]>
>> > Subject: [VOTE] Shall we release hbase-0.20.4 Release
>> Candidate 6 as  hbase-0.20.4?
>> > To: "HBase Dev List" <[EMAIL PROTECTED]>
>> > Date: Tuesday, May 4, 2010, 4:52 PM
>> > I've posted an hbase 0.20.4 release
>> > candidate 6 here:
>> >
>> >  http://people.apache.org/~stack/hbase-0.20.4-candidate-6/
>> >
>> > Should we release this candidate as hbase 0.20.4?
>>
>>
>>
>>
>>
>>
>
>
>
>
>

I've posted an hbase 0.20.5 release candidate 1 here:

  http://people.apache.org/~stack/hbase-0.20.5-candidate-1/

Should we release this candidate as hbase 0.20.5?

It has four bug fixes and it adds a few missing header licenses (See
CHANGES.txt or the release notes in the docs [1] for issues addressed
by this candidate).

Please take it for a test spin if you have a chance.  Vote closes
Thursday, the 20th of May.

Yours,
Your HBase Team

1. http://people.apache.org/~stack/hbase-0.20.5-candidate-1/hbase-0.20.5/docs/releasenotes.html