cocoon-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Sparkes <pe...@didm.co.uk>
Subject Re: PDF Image Caching Problem
Date Mon, 05 May 2008 15:40:53 GMT
Hi Mika,

Thanks for the info on open files with the 'noncaching' option. My 
application was crashing every 4 to 5 days and I could not find the 
reason; I got round it by restarting tomcat/cocoon once a day using a 
cron job. I think that I will also leave the image caching problem for 
latter, I have spent enough time on it and restarting cocoon once a day 
means that the pdf catches up with the html within 24 hours which, while 
not ideal, is acceptable

Thank for you help

Peter
> Hi Peter,
>
> unfortunetely I couldn't solve the problem and as it isn't so crucial 
> at the moment, I have left it behind for now on.
>
> Actually there might be another problem (or bug) lying around if you 
> produce htmls' with 'noncaching' option. Cocoon seems to be leaving 
> files open when using html-serializer with noncaching option in your 
> pipeline. And eventually your application can crash because you exceed 
> the maximum open files that your OS tolerates. (Cocoon 2.1.11 / Centos 
> 4.6)
>
> mika
>
> Peter Sparkes kirjoitti:
>> Hi
>>
>> I have tried using  <map:pipeline type="noncaching"> and invalidate 
>> the PDF by changing the XSLT that produces it. Neither makes any 
>> difference.
>>
>> It appears that FOP is getting the image from its own cache which is 
>> not being updated by the new version of the image. Like you I get the 
>> new image when I restart tomcat and hence cocoon
>>
>> Have you managed to fix the problem?
>>
>> Peter
>>> Hi,
>>>
>>> this is what Joerg Heinicke wrote about the similar problem of mine:
>>>
>>> My guess is it's actually the PDF that is taken from the cache. I 
>>> guess the cache key for the PDF does not take included resources 
>>> like your images into account. If that's true you just have to 
>>> invalidate the PDF. I don't know how exactly your PDF pipeline looks 
>>> like but you might want to "touch" (in the Linux sense, changing the 
>>> file's timestamp) an XSLT or the input XML.
>>>
>>> (See: image caching on 4-5 March at the users-list)
>>>
>>> mika
>>>
>>> Peter Sparkes kirjoitti:
>>>> Hi,
>>>>
>>>> I generate html and pdf, which contain images, from the a database. 
>>>> When I update an image the html correctly displays the new image, 
>>>> however the pdf incorrectly contains the old image. If I then 
>>>> change the new image file name the pdf displays the correct image. 
>>>> When I change the image file name back the old image is displayed 
>>>> in the pdf, not the new image.
>>>>
>>>> It appears that the pdf is using a cached version of the image not 
>>>> the new version. Please, how can I get the pdf to display the new 
>>>> image rather than the old cashed image. I can't change the image 
>>>> file name in the production system.
>>>>
>>>> I am using Cocoon 2.1.10
>>>>
>>>> Thank you
>>>>
>>>> Peter Sparkes

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


Mime
View raw message