The issue was identified and fixed.
https://issues.apache.org/jira/browse/COUCHDB-926 is now closed.

The same type of issue happened with continuous replications and the
the authentication cache (_users database).
Both fixed in trunk, 1.1.x and 1.0.x.

1.0.2 and 1.1.0 will contain the fixes.



On Tue, Oct 26, 2010 at 2:18 PM, Adam Kocoloski <[email protected]> wrote:
> I think that's the right approach.  Thanks for the JIRA ticket.
>
> Adam
>
> On Oct 26, 2010, at 3:10 AM, Russell wrote:
>
>> Thanks guys. I think I'm going to log it anyway, this problem is too
>> serious to just hope for the best.
>>
>> On Mon, Oct 25, 2010 at 10:21 PM, Filipe David Manana
>> <[email protected]> wrote:
>>> On Mon, Oct 25, 2010 at 5:01 PM, Russell <[email protected]> wrote:
>>>> So should I bother creating a bug in the issue tracker or can we
>>>> assume it will be fixed in the next version?
>>>
>>> I think it's only worth if the issue doesn't seem to be fixed on 1.0.x.
>>> 1.0.2 is going to be released soon (days, weeks? I can't tell), which
>>> is based on branch 1.0.x.
>>>
>>>>
>>>> On Mon, Oct 25, 2010 at 1:08 PM, Filipe David Manana
>>>> <[email protected]> wrote:
>>>>> The commit Chris did (2nd October, 1.0.x) also helps here:
>>>>>
>>>>> http://svn.apache.org/viewvc?view=revision&revision=1003718
>>>>>
>>>>> On Mon, Oct 25, 2010 at 9:16 AM, Christian Scharr
>>>>> <[email protected]> wrote:
>>>>>> As I can see in the dev-list, a possible way to fix this has been posted 
>>>>>> on Sept. 25 by Filipe
>>>>>>
>>>>>> -----
>>>>>> Quote:
>>>>>> -----
>>>>>> "Filipe,
>>>>>>
>>>>>> Thanks very much for looking at this, I'll get one of our guys to check 
>>>>>> out the code and put it on one of our servers, we can normally tell 
>>>>>> pretty quickly if its working or not.
>>>>>>
>>>>>> Will let you know the results
>>>>>>
>>>>>> Stephen"
>>>>>> -----
>>>>>>
>>>>>> But I can't find any further informations regarding the final fix of 
>>>>>> this bug.
>>>>>> Maybe it's targeted for 1.0.2 ...
>>>>>>
>>>>>> Mit freundlichen Grüßen,
>>>>>> Greets,
>>>>>>
>>>>>> Christian Scharr
>>>>>> Dipl. Wirtschaftsinformatiker (BA)
>>>>>> MESO-/InHouse-Entwicklung
>>>>>> ----------
>>>>>> HSH Soft- und Hardware Vertriebs GmbH
>>>>>> Rudolf-Diesel-Straße 2
>>>>>> 16356 Ahrensfelde
>>>>>> Tel. (Zentrale): +49 (0)30/94 004 0
>>>>>> Tel. (Hotline): +49 (0)30/94 004 444
>>>>>> Fax: +49 (0)30/94 004 400
>>>>>> eMail: [email protected]
>>>>>> ---
>>>>>> Amtsgericht Frankfurt (Oder): HRB 7352 FF
>>>>>> Geschäftsführer: Stephan Hauber
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: Russell [mailto:[email protected]]
>>>>>>> Sent: Monday, October 25, 2010 10:04 AM
>>>>>>> To: [email protected]
>>>>>>> Subject: Compaction and unreleased file descriptors
>>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> I'm currently battling with 1.0.1 production servers that require the
>>>>>>> CouchDB service to be restarted every couple of days due to the fact
>>>>>>> that CouchDB does not release the file descriptors of compacted
>>>>>>> databases. As a result the free space on my servers simply disappear
>>>>>>> over a couple of days, causing the servers to crash unless I stop and
>>>>>>> restart the CouchDB servers (at which point the release of the file
>>>>>>> descriptors returns the free space to the system).
>>>>>>>
>>>>>>> Having searched the mailing list, I came upon a thread titled "Couch
>>>>>>> not releasing deleted files" in september that seems to indicate that
>>>>>>> the couch devs are aware of the issue, however I cannot find a bug in
>>>>>>> the issue tracker relating to the problem.
>>>>>>>
>>>>>>> Does anyone know what the status of the issue is at the moment?
>>>>>>>
>>>>>>> Regards,
>>>>>>>  Russell
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Filipe David Manana,
>>>>> [email protected], [email protected]
>>>>>
>>>>> "Reasonable men adapt themselves to the world.
>>>>>  Unreasonable men adapt the world to themselves.
>>>>>  That's why all progress depends on unreasonable men."
>>>>>
>>>>
>>>
>>>
>>>
>>> --
>>> Filipe David Manana,
>>> [email protected], [email protected]
>>>
>>> "Reasonable men adapt themselves to the world.
>>>  Unreasonable men adapt the world to themselves.
>>>  That's why all progress depends on unreasonable men."
>>>
>
>



-- 
Filipe David Manana,
[email protected], [email protected]

"Reasonable men adapt themselves to the world.
 Unreasonable men adapt the world to themselves.
 That's why all progress depends on unreasonable men."

Reply via email to