Hi
Thanks for providing the log file!
My assumption was correct. The index data within the
{name}.solrindex.zip file gets ignored and because of that the
initialization fails. Here the according sections in the log
03.02.2013 20:50:44.626 *DEBUG* [Thread-40]
org.apache.stanbol.commons.solr.managed.impl.ManagedSolrServerImpl >
start initializing SolrIndex {}<sitename>
03.02.2013 20:50:44.627 *WARN* [Thread-40]
org.apache.stanbol.commons.solr.utils.ConfigUtils Context <sitename>/
not found in resource <sitename>\conf\admin-extra.html -> ignored!
[..a lot more ignored files..]
resulting in the
03.02.2013 20:50:44.652 *ERROR* [Thread-40]
org.apache.stanbol.commons.solr.managed.impl.ManagedSolrServerImpl
Exception while activating Index 'default:<sitename>'!
java.lang.IllegalArgumentException: The parsed File
'<root>\stanbol\indexes\default\<sitename>-2013.02.03' MUST represent
a Directory!
So if - as you have confirmed - the {sitename} of the solrindex.zip
file is in fact the same as the {sitename} within the archive i am now
assuming that this has todo with the sitename you are using. As I
assume that you have modified the log to use '<sitename>' instead of
the real sitename there is not much I can do other than asking if you
use some non ASCII chars in your name. Because this would suggest that
this issue is related to character encoding issues with zip archive
entries.
Can you please confirm this assumption. If true I would also expect
that choosing an ASCII only site name should solve this issue.
best
Rupert
On Mon, Feb 4, 2013 at 6:23 PM, Alexey Kudinov <[email protected]> wrote:
> please find the debug log attached
> regards,
> alexey
>
> ---------- Forwarded message ----------
> From: "Alexey Kudinov" <[email protected]>
> Date: Feb 3, 2013 8:56 PM
> Subject: RE: errors deploying a cached referenced site
> To: <[email protected]>
>
> Hi Rupert,
> Please find the log attached.
>
> -----Original Message-----
> From: Rupert Westenthaler [mailto:[email protected]]
> Sent: Sunday, February 03, 2013 11:37 AM
> To: [email protected]
> Subject: Re: errors deploying a cached referenced site
>
> Hi
>
>
> On Sun, Feb 3, 2013 at 9:48 AM, Alexey Kudinov <[email protected]>
> wrote:
>> I haven't changed any file name prior installation. I followed exactly
>> the tutorial on site. i gave a meaningful name to the site in
>> configuration file before indexing and that's it. somehow the directory
>> wasn't created.
>>
>
> Ok, but than I have no clue why the directory is not created. Can you please
> start the Stanbol Server with log level DEBUG (-l DEBUG), reproduce this
> error and provide the resulting log file.
>
> best
> Rupert
>
>> Regards,
>> alexey
>> On Feb 3, 2013 9:06 AM, "Rupert Westenthaler"
>> <[email protected]>
>> wrote:
>>
>>> Hi Alexey
>>>
>>> On Sat, Feb 2, 2013 at 11:06 PM, Alexey Kudinov
>>> <[email protected]>
>>> wrote:
>>> > I'm getting the following errors after deployment:
>>> >
>>> >
>>> >
>>> > 02.02.2013 23:27:45.932 *ERROR* [Thread-43]
>>> > org.apache.stanbol.commons.solr.managed.impl.ManagedSolrServerImpl
>>> Exception
>>> > while activating Index 'default:<..sitename..>'!
>>> > java.lang.IllegalArgumentException: The parsed File
>>> > '<..root..>\stanbol\indexes\default\<..sitename..>-2013.02.02' MUST
>>> > represent a Directory!
>>> >
>>> > at
>>> >
>>> org.apache.stanbol.commons.solr.SolrServerAdapter$SolrCoreProperties.
>>> setCore
>>> > Dir(SolrServerAdapter.java:856)
>>> >
>>> > at
>>> >
>>> org.apache.stanbol.commons.solr.managed.impl.ManagedSolrServerImpl.ac
>>> tivateC
>>> > ore(ManagedSolrServerImpl.java:839)
>>> >
>>> > at
>>> >
>>> org.apache.stanbol.commons.solr.managed.impl.ManagedSolrServerImpl.up
>>> dateCor
>>> > e(ManagedSolrServerImpl.java:788)
>>> >
>>> > at
>>> >
>>> org.apache.stanbol.commons.solr.managed.impl.ManagedSolrServerImpl$In
>>> dexUpda
>>> > teDaemon.run(ManagedSolrServerImpl.java:1163)
>>>
>>> The stacktrace suggest that the copying of the data form the Archive
>>> to the target directory was completed successful (because this part
>>> of the code is done before the location that throws the excaption).
>>> But than the check if the target directory exists fails. A detailed
>>> look at the code revealed that this can only happen if not a single
>>> file of the {sitename}.solrindex.zip file was copied (because
>>> directories are created lazily if actual resources are copied).
>>>
>>> Because of that I assume that you have renamed the
>>> "{sitename}.solrindex.zip" so that the {sitename} in the filename
>>> does no longer match the root directory within the archive. However
>>> this is required so that the initialization process can determine the
>>> relative path for copying the files (see specification in the section
>>> "Managing Solr Indexes" [1] of the commons-solr documentation).
>>>
>>> Users that want to rename a "{sitename}.solrindex.zip" file need
>>> therefore to extract the archive, change the name of the root
>>> directory and than re-comress the data.
>>>
>>> If this is indeed the case than the thrown exception is not very
>>> helpful and I will need to improve error handling to indicate the
>>> actual cause of the Problem.
>>>
>>> best
>>> Rupert
>>>
>>> [1]
>>> http://stanbol.apache.org/docs/trunk/utils/commons-solr#managing-solr
>>> -indexes
>>>
>>> --
>>> | Rupert Westenthaler [email protected]
>>> | Bodenlehenstraße 11 ++43-699-11108907
>>> | A-5500 Bischofshofen
>>>
>
>
>
> --
> | Rupert Westenthaler [email protected]
> | Bodenlehenstraße 11 ++43-699-11108907
> | A-5500 Bischofshofen
--
| Rupert Westenthaler [email protected]
| Bodenlehenstraße 11 ++43-699-11108907
| A-5500 Bischofshofen
