FirebirdSQL / firebird

Firebird server, client and tools
https://www.firebirdsql.org/
1.21k stars 209 forks source link

Replication problem #7575

Open Dany40 opened 1 year ago

Dany40 commented 1 year ago

Hello,

Using Firebird 4.0.2. My replication configuration is:

database = C:\DATOS\00498829.FDB { journal_directory = C:\JOURNAL\00498829\ journal_archive_command = "C:\FTPSend $(pathname)" journal_segment_count = 0 journal_archive_timeout = 10 }

If FTPSend returns always 1, Firebird is not creating more than 8 files in the journal folder; and on the replication.log, it is writting:

MyPC (primary) Mon May 8 08:45:47 2023 Database: C:\DATOS\00498829.FDB WARNING: Out of available space in journal segments, waiting for archiving...

And on the firebird.log it is writting:

MyPC Mon May 8 08:45:47 2023

Error in replicator thread
Out of available space in journal segments

And if I try to generate more changes on the database, if finally crash.

Regards.

lwennekers commented 2 months ago

Hello,

On Firebird 4.0.2 I've seen the same issue. After a while I got some other errors too (see below) perhaps it helps. It was not clear if the replication was slow or stopped.

P291ST201-1 Mon Apr 15 10:37:09 2024 Error in replicator thread Out of available space in journal segments

P291ST201-1 Tue Apr 16 01:17:14 2024 Operating system call ReleaseSemaphore failed. Error code 6

P291ST201-1 Tue Apr 16 02:10:26 2024 Operating system call SetFilePointer failed. Error code 112

P291ST201-1 Tue Apr 16 02:10:26 2024 TPC: Cannot initialize the shared memory region (header) operating system directive SetFilePointer failed Onvoldoende schijfruimte beschikbaar.

aafemt commented 2 months ago

Windows error 112: ERROR_DISK_FULL There is not enough space on the disk.

It looks like problem with your hardware, not Firebird.

lwennekers commented 2 months ago

I understand. But why are the replication journals remaining? They are the cause of a full disk

lwennekers commented 2 months ago

On issue #7575 I added some information also about the number of files increasing during replication because of replication logging. Could it be related? It looks like a replication problem