lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From MakMak <>
Subject Re: readModifiedUTF8String stuck
Date Tue, 21 Apr 2009 22:25:24 GMT


   Ran CheckIndex. This is what it prints out:

cantOpenSegments: false
numBadSegments: 0
numSegments: 14
segmentFormat: FORMAT_HAS_PROX [Lucene 2.4]
segmentsFileName: segments_2od
totLoseDocCount: 0
clean: true
toolOutOfDate: false

So I guess everything is fine ..! 

Our application is a very resource intensive application. We have everything
in the RAM for best possible response time and granularity of our
authorization model. We give around 6GB of RAM for the application. During
stress tests the RAM consumption gets used up to "2% free". But this has
always been the case when we were on 2.3.2. With 2.4.1 is when we started
hitting upon these stuck threads.

There are other differences between the customer and our setup apart from
SAN like http/https, webservices enabled/disabled etc but these shouldnt
have anything to do with this problem.

Currently we are trying to bring the indexes  out of the SAN and into the
local filesystem to see if these issues remain. Will keep you posted. 

Anymore hints?

-thanks a ton for all the help

Michael McCandless-2 wrote:
> On Mon, Apr 20, 2009 at 6:07 PM, MakMak <> wrote:
>>   I made a standalone tool like you suggested which prints out the size
>> of
>> each doc in the index, none of the docs are more than 1MB !!! The queries
>> are the same. They repeat throughout the test. We give about 6GB of heap
>> to
>> the application and yes we are on 64 bit JVM.
> OK.
> Any luck running CheckIndex?  Can you run w/ verbose GC, to see if GC
> is running...?
>> I hit upon another such issue here
>> Do you think this may be related? There was no resolution on this thread
>> though. We only upgraded lucene version and I am absolutely certain
>> nothing
>> else changed.
> They could be related.  They sound very similar.
>> One more thing, we tried to reproduce the customer environment here
>> exactly
>> like they have it there. The only difference is, the index on the
>> customer's
>> end resides in a SAN.
> Are you not able to repro the issue locally?  So the SAN is the only
> known difference?
> Mike
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

View this message in context:
Sent from the Lucene - Java Users mailing list archive at

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message