pdfbox-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Tilman Hausherr <THaush...@t-online.de>
Subject Re: PDFBox loading larger PDF
Date Sat, 28 Mar 2015 08:59:29 GMT
Hi,

Sorry, no, we need the whole file.

You could try to use a scratch file in load() or (better) loadNonSeq.

I also see my last mail was a bit unclear
- with "merging a file with itself" I meant doing this until you get a 
huge non confidential file. Try e.g. with the PDF specification file.
- with "ab.exe" I meant "apache benchmark". It isn't part of the tomcat 
distribution, but one can extract it from some apache webserver 
distributions.

I assume you use the 1.8 version. In the unreleased 2.0 version there 
would be a trick (with a few lines of code change) to save memory by not 
keeping the images unless needed.

Tilman

Am 28.03.2015 um 08:14 schrieb Swapnil Raverkar:
> Hi Tilman,
>
> For PDFBox is there any way to load the file in chunks instead of loading
> entire file in memory for populating the document object? It is consuming
> around 1GB heap memory at start-up for a single file (200 MB) for a single
> user and drops down to 600-700 MB during the processing time.
>
> Thanks,
> Swapnil
>
> On 27 March 2015 at 23:44, Tilman Hausherr <THausherr@t-online.de> wrote:
>
>> Errors that happen in concurrent situations are notoriously hard to find.
>> The best would be to prepare
>> - a large file of the kind you mention
>> - code of a servlet to be run in tomcat (smallest possible code that
>> brings the error)
>> then use ab.exe to stress-run the servlet in the tomcat 8 and fine tune
>> the parameters until it happens for sure, then send us all that.
>>
>> If the file is confidential, try if the error happens by merging a non
>> confidential file with itself.
>>
>> Tilman
>>
>> Am 27.03.2015 um 18:42 schrieb Swapnil Raverkar:
>>
>>   No this issue is not happening for a large file with a single active user.
>>>
>>> Thanks,
>>> Swapnil
>>>
>>> On 27 March 2015 at 01:34, Tilman Hausherr <THausherr@t-online.de> wrote:
>>>
>>>   Am 27.03.2015 um 08:49 schrieb Swapnil Raverkar:
>>>>   While loading larger PDF files more than 200 MB for 10 concurrent users
>>>>> with 3GB heap Space configured in JVM, we are getting following
>>>>> exception
>>>>> :
>>>>>
>>>>>   Does this also happen with a large file when only 1 user is active?
>>>>
>>>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
>> For additional commands, e-mail: users-help@pdfbox.apache.org
>>
>>


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: users-help@pdfbox.apache.org


Mime
View raw message