Extracting data upon failure

[Sugarat]

Hi All,

I\'m new to OpenKM and have a question. I just started using another DMS but it strangely died out of the blue. I\'d like to switch to OpenKM but my concern is how to retrieve files should OpenKM ever go wrong. I\'m not that fussed about the file metadata, but should I need to get to the uploaded files in a situation where OpenKM was broken, how could I do so?

It looks like the files are kind of \'sucked\' into OpenKM and cannot be found on the servers local filesystem?

Any tips to ease my nervousness about putting loads of info into OpenKM?

Thanks,

Adam

[jllort]

Don\'t worry.

Files on OpenKM can be stored on several ways, dbms ( postgre, mysql, oracle ) or file system.

By default The OpenKM configuration save files on file system under Jboss/repository, but are saved on a special structure - refer to jackrabbit if you want more information about it.

One think that always we\'ve got present since starting OpenKM is that end user is owner of files, and if want to move to other system this must be done easilly, we think it\'s some to evaluate comparting to other dms.

You must not try to recover - find - files from this structure (it\'s special blobs structure not tipical to facilite fast access and other improvements). On OpenKM there\'s a way to export all data structure to normal server file system. Connect as system user, go to admin menu options, an select export option.

Take a look at this threath about exporting option.

[Sugarat]

Hi,

I\'m not really sure how you can say \'Dont worry\' because OpenKM can export all files. Thats not much good if OpenKM is borked and won\'t start for some reason. I won\'t have the expertise to start it up again, which is why having the files stored as binary blobs makes me nervous. If OpenKM is dead then I\'ve lost all the files?

[jllort]

See files names on file system not ensures that if your computer corrupts you recover it. It\'s not a problem about how files are stored, is a problem if you don\'t make backups of OpenKM or other file system, database etc...

For example if your dms stores on a DBMS could not be so easy to know a direct correspondence about what you see on front and how is stored information on dbms. If you backup correctly database you don\'t lost information if your server corrupts that\'s the same idea.

You must ensure you backup well OpenKM like other system, and we recomend of course on production environment to have a UPS.

I talk about export, because some dms, especially some comertial ones, makes difficult to export files from theirs repositories, why ?

About OpenKM backup, if you install as you download, without modifying repository.xml parameters you must ensure to backup /Jboss/repository directory

[blandobardo]

I had properly backed up my repository, now I reinstalled OpenKm 2.0 because Openkm 3.0 doesn\'t have the same repository and the conversion from one to another hangs at 70% after more than 4 hours of work..
So I finally decided to reinstall the old OpenKm2.0 and hoped all my data were not lost but after endless fidging with configuration and so on I get a:

\"

Code: Select all

15:24:38,850 ERROR [DirectRepositoryModule] Cannot instantiate persistence manager org.apache.jackrabbit.core.persistence.bundle.DerbyPersistenceManager: \'NODE_ID_HI\' is not a column in table or VTI \'APP.VERSION_REFS\'.
javax.jcr.RepositoryException: Cannot instantiate persistence manager org.apache.jackrabbit.core.persistence.bundle.DerbyPersistenceManager: \'NODE_ID_HI\' is not a column in table or VTI \'APP.VERSION_REFS\'.: \'NODE_ID_HI\' is not a column in table or VTI \'APP.VERSION_REFS\'.

\"
all my files now are binary and useless and there\'s no exporting without making openkm2.0 work.. If I try to login I get a \"The repository home /media/raid_data/repository appears to be already locked by the current process.\"

I have opened a separate forum on this issue at http://www.openkm.com/Usage/1639-Lost-m ... .html#1639 this is just to reopen this thread..

[pavila]

Please, do not mix issues in the same thread.

[pavila]

If you make regular backups you wont lost information. OpenKM is an Open Source project so it will remain available for everyone who wants to use it. Also you can get the source code, and this is a big advantage because you can enhance it if our company dissapears. AOpenKM also is based on Jackrabbit, an implementation of JSR-170 so the repository structure is easily manipulated. No obscure algorithms or propietary data structure used, so I don\'t see where is the problem.

[pavila]

You can also contact us for commercial commercial and help us to improve OpenKM and make a better product.