directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Selcuk AYA <ayasel...@gmail.com>
Subject Re: Release troubles and failing tests
Date Wed, 09 May 2012 20:39:06 GMT
What I saw happening was there is a reader at version 188. Some reader
somehow does not decremented the ref count at this version  and
writers keep bumping up the version upto until ~600. Then we run out
of freeable cache elements.

We always have problem with the same version(188) and the problem
always happens when updateding the nbchildren nbdescendants of the
parents in rdn index. This is a new code but I dont see anything wrong
with it. What I am thinking is that this new code exposed some bug in
LRUCache that was there before.

I am stuck at this point because of my ignorance of the build
environment. I use the debugger to step through the code and I have
the apacheds-jdbm as a module but the runtime uses a jdbm jar in
repository rather than the code I modify so I cannot add debug code to
jdbm. I really would appreciate some help here.

thanks
Selcuk
.
On Wed, May 9, 2012 at 10:40 AM, Emmanuel Lécharny <elecharny@gmail.com> wrote:
> Le 5/9/12 5:34 PM, Emmanuel Lécharny a écrit :
>
>> Le 5/9/12 6:47 AM, Emmanuel Lécharny a écrit :
>>>
>>>
>>>> I'll run the tests and let you know if it's any better.
>>>
>>>
>>> Still having issues, even after having closed the cursors in the
>>> PagedSearch :/
>>
>>
>> Ok, one step farther...
>>
>> I was able to reprduce the priblem, with some extra logs.
>>
>> When we get blocked, we have a full LRUCache :
>>
>> LRUCache:
>> (numEntries:1024,maxEntries:1024,cacheGets:8071,cacheMisses:35,cachePuts:1462,cachePutSleeps:57)
>>
>> but as I'm dumping the cache content just after, I can see that it
>> contains only 10 elements...
>>
>> Here, there are two possibilities
>> 1- the dump is wrong, and does not dump what is in the cache, but what is
>> in the cache after the blockage, and after some elements have been released
>> 2- we have some serious issue with the counters
>>
>> I do think that -2- is the problem. Will investigate more.
>
> Well, counters are just fine. So there must be somethong wrong elsewhere...
>
> Time for me to seriously check the LRUCache code, in order to understand how
> it works ! We have a cool tool at Apache to share some review :
> https://reviews.apache.org/dashboard/
> We could use it to comment the code, Selcuk.
>
> Sorry if I haven't spent as much time as needed on this part of the code
> today, I had to deal with some other things. Will be back later tonite.
>
> Thanks !
>
>
>
> --
> Regards,
> Cordialement,
> Emmanuel Lécharny
> www.iktek.com
>

Mime
View raw message