The configuration page of the Ark module isn’t loading either.
The BerkeleyDB is present, albeit somewhat big, ./omeka-s-files/arkandnoid/NOID/noid.bdb is about 356MB.
@Daniel_KM Could the database be corrupt? If so, any pointers on how I can fix the database? Is deleting the noid.bdb an option (keep in mind I already have 900K Arks minted with NAAN 60537). Is there a migration path from BerkeleyDB to MariaDB available?
However, when saving an Item there’s an error in the log (so did the ‘repair’ above succeed / was this the right way to repair???) which seems tot indicate that $ark = \Noid::mint($noid, $contact) failed:
TypeError: strlen(): Argument #1 ($string) must be of type string, null given in [..]/Ark/src/Name/Plugin/Noid.php:71
Stack trace:
#0 [..]/Ark/src/Name/Plugin/Noid.php(71): strlen()
#1 [..]/Ark/src/ArkManager.php(391): Ark\Name\Plugin\Noid->create()
#2 [..]/Ark/Module.php(324): Ark\ArkManager->createName()
#3 [..]/Ark/Module.php(282): Ark\Module->addArk()
So I still can’t save items with arks like I used to…
As a solution to my “corrupted noid database” I have deleted the noid.dbd file and created a new one (via the module configuration) where I changed the identifier template to be one alphanumeric character (l) longer. This way there won’t be duplicates with the already 900K published ARK identifiers.
I have no idea how the BerkeleyDB got corrupted and am disappointed that is seemed to be beyond repair. I hope the ARK module will make use of MariaDB in the future, noidphp already seems ready. This would mean no dependency on BerkeleyDB, but an already in use MariaDB. Of course this would mean, the module should also provide a migration option (from BerkeleyDB to MariaDB).