Closed gdower closed 6 months ago
This is very weird. Im investigating...
btw, the logs also say this:
Rematched 60 SpeciesEstimates from project 3. updated: 1, broken: 59
There was a bug in the code that swallowed some exceptions and converted them into interrupted exceptions, used to cancel operations. So there is some other, real exception underneath this problem that we cannot see so far. I will deploy the fix in an hour and try to sync millibase again, so we can hopefully see and ultimately fix the real problem
deployed. Finally we can ee a NullPointerException
Caused by: java.lang.NullPointerException: null at life.catalogue.assembly.TreeBaseHandler.processEnd(TreeBaseHandler.java:169) at life.catalogue.assembly.TreeCopyHandler.process(TreeCopyHandler.java:164) at life.catalogue.assembly.TreeCopyHandler.processLast(TreeCopyHandler.java:125) at life.catalogue.assembly.TreeCopyHandler.acceptThrows(TreeCopyHandler.java:115)
It was introduced by the sic
parsing in authorships. Datasets with sic
synonyms currently fail. I might need a day to fix this, its a bit more complex I am afraid.
deployed to prod just now and synced millibase successfully
@mdoering, when you cancelled the sector syncs over the [sic] bug, it seems to have broken the sync queue. It's not possible to re-sync these four datasets with sectors in cancelled state:
MilliBase, Mollusca, Isopoda, Bryozoa
I get a 201 on the sync POST, but it doesn't seem to sync. The other WoRMS sectors with syncs that weren't in cancelled state seemed to have worked.
Logs&_a=(columns:!(level,sectorKey,service,logger_name,message),filters:!()))