commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mark Thomas <ma...@apache.org>
Subject Re: DBCP deadlock. Is this bug DBCP-270?
Date Thu, 23 Jul 2009 10:02:06 GMT
Ying Chyn Ng wrote:
> Hi Mark,
> 
> Attached is the full thread dump of the DBCP deadlock issue I encountered.
> Is the information sufficient?  Please let me know what other info is
> required and how to collect it.

<snip/>

Yes, that was sufficient. Actually it was more than sufficient. All that was
required was the complete stack traces from the deadlock section.

Answers to your original questions below.


> ----- Original Message ----- From: "Mark Thomas" <markt@apache.org>
> To: "Commons Developers List" <dev@commons.apache.org>
> Sent: Wednesday, July 22, 2009 11:09 PM
> Subject: Re: DBCP deadlock. Is this bug DBCP-270?
> 
> 
>> Ying Chyn Ng wrote:
>>> Hello.
>>>
>>> I encountered an issue that looks similar to Bug DBCP-270.
>>> I wish to know:
>>> 1) Is it the same bug or a new one?
>> Don't know. Without the full stack traces for the locked threads it is
>> impossible to tell.

Yes, that is DBCP-270

>>> 2) Which patch can I apply and how to apply it?
>> That depends on what problem you are seeing.

You'd need to apply the patch for DBCP-270 to a local copy of DBCP and then
build Tomcat with your modified DBCP sources.

>>> 3) Will setting testWhileIdle="false" avoid the problem?

No. You'd need to completely disable the evictor thread. Note that there are
other potential deadlocks not related to the evictor that you may still hit.

If you have any more questions, please use either the Commons or Tomcat users
list as appropriate.

Mark

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message