On 4/30/2010 12:11 PM, Steve Singer wrote:
> Leon Mergen wrote:
>> Hello Steve,
>> 
>> On Fri, Apr 30, 2010 at 5:39 PM, Steve Singer <ssinger at ca.afilias.info> wrote:
>>>> I will try to find out what log entry is causing the trouble.
>>>>
>>> Do the tables that these errors come from typically get large insert/update
>>> statements or small ones?
>>>
>>> (this is directed at anyone who's seen the issue)
>> 
>> I'm still trying to isolate the issue a bit, but at the moment it
>> looks like this issue is triggered by a very large INSERT statement
>> (as in, a statement with a very large value);
>> 
>> Regards,
>> 
> 
> Your issue with the larger values are being caused by an attempt at 
> removing a memory leak checked into 2.0.3 that free's memory a bit too 
> soon.  I will post a patch for this later today.   I think the INSERT 
> statement is being corrupted causing invalid characters to be put in.
> 
> 
> What Devrim's is seeing where the initial copy fails would be a 
> different issue.

I understood Devrim as that it happens during the initial sync. That is 
not the copy_set.


Jan

-- 
Anyone who trades liberty for security deserves neither
liberty nor security. -- Benjamin Franklin