Mailing-List: contact couchdb-user-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: couchdb-user@incubator.apache.org
Received-SPF: neutral (athena.apache.org: local policy)
Message-Id: <2FFA35F8-612E-4624-99B7-F090D9E66045@apache.org>
From: Damien Katz <damien@apache.org>
To: couchdb-user@incubator.apache.org
In-Reply-To: <7db9abd30809301752y6eae1cdfy623f5becee26a06@mail.gmail.com>
Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes
Content-Transfer-Encoding: 7bit
Mime-Version: 1.0 (Apple Message framework v929.2)
Subject: Re: Exception during compaction
Date: Tue, 30 Sep 2008 21:34:23 -0400
References: <7db9abd30809292010m37deeabav9611c9eceda3a0fe@mail.gmail.com>
 <7920B692-4682-4573-B7EB-6A2DB784B7C7@apache.org>
 <7db9abd30809301752y6eae1cdfy623f5becee26a06@mail.gmail.com>

I've tracked this down, it looks like I fixed this (inadvertently) in  
trunk, but there is still a small bug to be fixed (doesn't cause an  
error, just extra index data kept around). If you switch to trunk you  
won't have this problem. If needed we can fix this in a 0.8.2 release.

-Damien


On Sep 30, 2008, at 8:52 PM, kowsik wrote:

> I should add that for the fun of it I ran a compact during a
> write-heavy phase and then stopped the writes. I noticed a message
> indicating compaction was suspended (not the exact words) and then it
> was successful. Remember reading something about this in the wiki that
> compaction could get starved if there too many writes on the db.
>
> This was the only unusual step that I remember. After this happened,
> even if there were no more active writes to the db, compaction always
> fails with this exception.
>
> Dunno if that helps.
>
> K.
>
> On Tue, Sep 30, 2008 at 9:17 AM, Damien Katz <damien@apache.org>  
> wrote:
>> I've seen several reports something like this in 0.8.1. There  
>> definitely
>> looks to be a bug here. Has anyone seen anything like this in trunk?
>>
>> kowsik, do you have any reproducible steps to make this happen? If  
>> not, do
>> you know if the database had already been compacted at least once  
>> before
>> this happened and was the database created with the same version?
>>
>> -Damien
>>
>>
>> On Sep 29, 2008, at 11:10 PM, kowsik wrote:
>>
>>> Anyone seen this? Running couchdb 0.8.1 with erlang @R12B-3_2 on
>>> Darwin 10.4.11. Happens when I compact the database. There are no
>>> writers active at this point. Just viewing the existing data.
>>>
>>> Thanks,
>>>
>>> K.
>>>
>>> =ERROR REPORT==== 29-Sep-2008::18:30:06 ===
>>> Error in process <0.82.0> with exit value:
>>>
>>> {function_clause,[{couch_db,'-copy_docs/3-fun-0-', 
>>> [not_found,<0.58.0>,{stream,<0.84.0>,<0.83.0>},<0.83.0>]}, 
>>> {lists,map,2},{lists,map,2},{couch_db,copy_docs,3}, 
>>> {couch_db,copy_compact_docs,2},{couch_db,start_copy_compact_int,2}]}
>>>
>>> [error] [<0.47.0>] {error_report,<0.22.0>,
>>>  {<0.47.0>,supervisor_report,
>>>   [{supervisor,{local,couch_server_sup}},
>>>    {errorContext,child_terminated},
>>>    {reason,
>>>        {function_clause,
>>>            [{couch_db,'-copy_docs/3-fun-0-',
>>>
>>> [not_found,<0.58.0>,{stream,<0.84.0>,<0.83.0>},<0.83.0>]},
>>>             {lists,map,2},
>>>             {lists,map,2},
>>>             {couch_db,copy_docs,3},
>>>             {couch_db,copy_compact_docs,2},
>>>             {couch_db,start_copy_compact_int,2}]}},
>>>    {offender,
>>>        [{pid,<0.59.0>},
>>>         {name,"blah"},
>>>         {mfa,
>>>             {couch_db,open,
>>>                 ["blah","/usr/local/var/lib/couchdb/blah.couch"]}},
>>>         {restart_type,transient},
>>>         {shutdown,infinity},
>>>         {child_type,supervisor}]}]}}
>>
>>