You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Describe the bug
Nuking / Deleting / Moving / Wiping a release with active transfers might kill the slave process as some threads are writing/reading as the folder is being moved/removed on the backing device.
Autofreespace running before the slave has finished remerging would trigger this as well.
Expected behavior
Abort+block transfers to paths being nuked or wiped, similar to glftpd's "this file is being deleted" and don't crash the slave
Maybe we can be a bit smarter on how/when we raise exceptions for missing files, most of the time the backing filesystem will not be the problem.
thrownewFileNotFoundException(path + " wasn't found in any root");
}
DEBUG 18:41:07,397 [Slave Main Thread] org.drftpd.slave.Slave [] - Slave fetched org.drftpd.common.network.AsyncCommandArgument[index=f1,name=checksum,args=/linux/irunarch/btw.iso]
DEBUG 18:41:07,398 [AsyncCommandHandler - class org.drftpd.common.network.AsyncCommandArgument] org.drftpd.slave.Slave [] - Slave wrote response - AsyncResponseException exception is - /linux/irunarch/btw.iso wasn't found in any root
DEBUG 18:41:07,398 [AsyncCommandHandler - class org.drftpd.common.network.AsyncCommandArgument] org.drftpd.slave.Slave [] -
DEBUG 18:41:07,412 [Slave Main Thread] org.drftpd.slave.Slave [] - Lost connection to the master, may have been kicked offline
WARN 18:41:07,412 [Slave Main Thread] org.drftpd.slave.Slave [] - Shutdown() called
WARN 18:41:07,412 [Slave Main Thread] org.drftpd.slave.Slave [] - Closing _sin
WARN 18:41:07,412 [Slave Main Thread] org.drftpd.slave.Slave [] - Closing _sout
WARN 18:41:07,412 [Slave Main Thread] org.drftpd.slave.Slave [] - Closing _socket
INFO 18:41:07,412 [Slave Main Thread] org.drftpd.slave.Slave [] - Setting slave status online to: false
The text was updated successfully, but these errors were encountered:
Describe the bug
Nuking / Deleting / Moving / Wiping a release with active transfers might kill the slave process as some threads are writing/reading as the folder is being moved/removed on the backing device.
Autofreespace running before the slave has finished remerging would trigger this as well.
Expected behavior
Maybe we can be a bit smarter on how/when we raise exceptions for missing files, most of the time the backing filesystem will not be the problem.
Additional context
Example of a crash:
drftpd/src/core/slave/src/main/java/org/drftpd/slave/vfs/RootCollection.java
Lines 264 to 272 in 5a5ac84
The text was updated successfully, but these errors were encountered: